Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are provided, so the description must carry the full behavioral burden. 'View' (查看) implies a read-only listing, and the availability bracket usefully discloses role and platform gating, but the description says nothing about whether the pending items can be acted on, whether reads are safe, or what the queue contains.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

It is a single front-loaded sentence that puts the purpose before the bracketed metadata, with no filler. The bracketed availability block is dense but carries real information rather than padding.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a zero-parameter read tool with no output schema, the definition covers what the tool is but not what it returns: what a 'pending action' looks like, whether the list is empty in common cases, or what an agent should do with the result. Adequate but with clear gaps.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The tool takes zero parameters, so the baseline is 4; the description correctly implies no input is needed to fetch the queue. Nothing is missing on the parameter side.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description names a specific verb and resource: viewing the pending-action queue left over from older versions ('查看旧版本遗留的待执行动作队列'). The 'legacy compatibility interface' framing distinguishes it from active agent tools such as boss_agent_run, boss_agent_review, and boss_agent_stats, though it never names those siblings explicitly.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

Usage is only implied: 'compatibility interface' and 'left over from older versions' suggest it is for legacy workflow state, not for normal operation. The bracketed availability metadata (roles=candidate, recruiter; platforms=zhipin) gives a real precondition, but there is no explicit when-to-use versus when-not-to-use or a named alternative.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are provided, so the description carries the full disclosure burden. '查看' (view) plus the eligibility bracket establishes this as a read operation scoped to specific roles and platforms, which is useful. It still omits what the queue contains, whether it returns pending items with IDs suitable for approve/reject, and whether any state changes occur, which matters for a review-queue tool with action siblings.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

One compact sentence with the resource view front-loaded, followed by a clearly delimited availability block. Nothing is redundant, though the bracketed metadata is dense and machine-style rather than explanatory, which slightly reduces readability.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

With no parameters, no annotations, and no output schema, the description is the only source of truth. It establishes eligibility and that this is a legacy compatibility view, but does not describe what the agent receives back or how the review queue relates to the approve/reject siblings, leaving a meaningful gap for a workflow tool.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The input schema has zero parameters, so there is no parameter meaning for the description to add; the baseline for an empty parameter surface is 4. The bracketed role and platform constraints are not parameters but effectively communicate invocation eligibility, which is a small bonus rather than a deduction.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description gives a concrete verb and resource: a view of the legacy manual review queue exposed through the compatibility interface. It does not, however, distinguish this tool from closely related siblings such as boss_agent_pending or boss_agent_review_approve/reject, so an agent cannot infer from the description alone which of these queues or actions it should use.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracket supplies real preconditions: the allowed roles (candidate, recruiter) and platforms (zhipin only). That is a genuine eligibility gate, but the description never says when to pick this tool over boss_agent_pending, nor when not to use it (for example, only in legacy/compatibility flows), leaving the routing decision implicit.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are supplied, so the description carries the full burden. It does disclose one side effect — a write into the pending queue — but says nothing about what happens to the original review item, whether the operation is idempotent or reversible, what errors occur on a stale id, or required permissions.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

A single compact sentence with the effect front-loaded, followed by an availability qualifier that earns its place. No redundant restatement of the tool name beyond what is needed.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a one-parameter tool with full schema coverage and no output schema, the minimum is met. The gap is routing and outcome: an agent still cannot tell from the text whether this approves the item, how it differs from boss_agent_review, or what the pending-queue write produces.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Only one parameter, and schema coverage is 100% with the description "review item id" already in the schema. The description adds no format, provenance, or sourcing detail (e.g., where to obtain the id), so baseline 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

"处理旧版本复核项并写入 pending 队列" names a resource (legacy review items) and an effect (write to the pending queue), and the 'compatibility interface' framing signals it is a legacy shim. But the verb '处理' (process) is vague, and nothing says this is the 'approve' counterpart to boss_agent_review_reject / boss_agent_review, so an agent must infer the action from the name alone.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability block (roles=candidate/recruiter, zhipin platform) is a real gating rule and useful. However, there is no guidance on when to choose this over the modern boss_agent_review path or boss_agent_pending, which is the critical routing decision for a compat shim.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description usefully discloses that the call changes a review item's state and records a skip event, while also supplying role and platform constraints. It does not explain whether rejection is reversible, how it affects the associated automation workflow, or what happens on failure.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

The single front-loaded statement communicates the purpose, logged side effect, and availability without filler. The bracketed availability block is dense but compact and useful.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a simple two-parameter operation, the action, target, side effect, availability, and schema provide minimum viable invocation context. With no annotations or output schema, the missing lifecycle consequences, ID-discovery workflow, and result behavior remain meaningful gaps.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, and both id and reason are already documented there, so the baseline score applies. The description adds no source for the review-item ID or guidance on when the optional reason should be supplied.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description gives a specific action and target—rejecting legacy review items—and adds that a skip event is recorded. It does not explicitly relate the operation to siblings such as boss_agent_review_approve or boss_agent_pending, so sibling differentiation remains incomplete.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The compatibility/legacy wording and explicit candidate, recruiter, and zhipin availability constraints establish a narrow usage context. However, there is no workflow guidance on when rejection is appropriate, how to obtain a review item, or when to use boss_agent_review_approve instead.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are supplied, so the description carries the full burden. It says 'execution' happens on a live recruitment platform, but never states side effects (messages or greetings sent to real candidates), whether the run can be interrupted or resumed, how it relates to the stopped state implied by the resume parameter, rate limits, or what happens on partial failure. The availability clause is useful metadata, but the risk profile of an autonomous agent run is left undisclosed.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

One compact sentence lists the pipeline stages up front, followed by a bracketed availability tag. Nothing is redundant, though listing five stages in a comma chain is slightly dense and the availability tag could be separated more clearly.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

With no output schema and no annotations, the description should say more about what a round produces (leads, decisions, stats) and when the run halts. Parameter documentation in the schema is thorough, which offsets some of the gap, but the behavioral picture for an autonomous automation tool remains thin.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, and the three parameters (limit, resume, dry_run) already carry detailed semantics in the schema, including the dry_run default of true and the strict confirmation needed for resume. The description adds no parameter meaning beyond that, so the baseline of 3 for fully documented schemas applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description names a concrete operation: a single automation round covering scan, decision, threshold control, execution and lead generation, along with the roles and platforms it applies to (zhipin). That is enough to distinguish it from granular siblings like boss_agent_stop, boss_agent_pending or boss_watch_run, though it never names those siblings explicitly.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability clause (roles=candidate/recruiter; platforms=zhipin) gives a clear precondition for use, which is genuinely useful. However, it offers no guidance on when to choose this over boss_watch_run, boss_agent_pending or boss_agent_stop, and no explicit when-not-to-use condition, so selection between it and its closest siblings must be inferred.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are supplied, so the description carries the full behavioral burden. It does add real context with the availability constraints (roles=candidate/recruiter; platforms=zhipin), which tells the agent who may call it. However, it never states that the call is read-only, whether it touches shared state, how fresh the counters are, or what the figures represent, which are the traits that matter most for a statistics endpoint.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

One front-loaded clause states the domain and the statistic categories, followed by a bracketed availability note. There is no filler, though the availability bracket is dense and mixes two different constraint families without explanation.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

There is no output schema, so the description is the only source of information about return values; it names four statistic categories but does not indicate the shape, units, or time window of any of them. Since this is a low-complexity zero-parameter read, the definition is adequate, but it could do more to describe what the caller will see.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The tool takes zero parameters, so per the baseline there is no parameter semantics to explain and the schema is exhaustive at 100% coverage. Nothing in the description is needed to call it, and nothing is missing.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description names a specific verb (view/查看) and enumerates the resources measured: execution, skipped, historical queue, and circuit-breaker statistics. That is clearly a reporting tool for the agent automation, distinguishable from the general boss_stats, though it does not explicitly say how it differs from boss_stats or boss_agent_pending.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

No when-to-use guidance is given. With many adjacent siblings (boss_stats, boss_agent_pending, boss_agent_run, boss_agent_review), the definition never says when to consult this stats view versus those, nor does it state any prerequisites or exclusions. The reader must infer that it is a read-only monitoring call.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are provided, so the description carries the full behavioral burden. It does disclose a meaningful trait — that this sets a breaker state affecting '后续自动执行' (subsequent automatic execution), suggesting persistent effect rather than a one-shot action — plus availability constraints (roles and platforms). It omits whether the state is reversible, whether in-progress executions are aborted, and what authority is required, which are important for a stop/kill-switch tool.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

The core action is front-loaded in one compact clause, with the availability matrix bracketed at the end. No redundant sentences. The bracketed availability block is somewhat dense and templated but does not obscure the primary statement.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a simple one-optional-parameter toggle with no output schema, the description covers what the tool does and who can call it. It is still incomplete on the operationally critical points: effect on currently running automation, reversibility, and post-stop state, which an agent needs before triggering a circuit breaker.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%: the single optional 'reason' parameter is documented in the schema with a default of 'manual-stop'. The description adds no additional meaning about how the reason is used or where it is recorded, so this sits at the baseline for high schema coverage.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description names a specific action on a specific resource: '打开熔断，停止后续自动执行' (open the circuit breaker, halt subsequent automatic execution) within recruiting automation. This is distinguishable from siblings like boss_agent_run or boss_agent_train. It stops short of 5 because it doesn't clarify the granularity — whether it kills in-flight work, cancels queued tasks, or just sets a persistent flag.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

'打开熔断' (open the circuit breaker) implies an emergency/pause scenario, which gives an agent a reasonable usage hint. However, there is no explicit statement of when to prefer this over simply not calling boss_agent_run, no mention of prerequisites, and no indication of how to resume automation afterward.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations supplied, the description carries the full burden. It does disclose one meaningful trait: actions meeting the threshold are rehearsed in dry-run by default, which signals a non-destructive default. It omits whether any state is persisted, whether a training run can be undone, and what the caller should expect back.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

It is a single compact line with the core mode stated first and availability bracketed at the end, so there is no wasted wording. The terseness, however, comes at the cost of the specifics an agent needs, so compactness here reads as under-specification rather than efficiency.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

The tool is simple on paper (zero required params, no output schema, no nested objects), so the description does not need to explain return values. Still, for an automation/calibration tool sitting among many agent_* siblings with no annotations, an agent cannot determine the outcome or scope of a training run from this text.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Only one optional parameter exists and its schema description ("本轮最多处理多少个会话") is fully documented at 100% coverage. The tool description adds nothing about 'limit', but with a single fully-described optional param the baseline of 4 is appropriate.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description frames this as "招聘自动化" in "训练校准模式" that by default dry-runs actions meeting a threshold, which conveys a calibration/training operation. However, it never says what is actually being trained (a model? a rule set?) or how this differs from the similarly named sibling boss_agent_run, so the agent must infer the distinction.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability line gives environment gating (roles=candidate/recruiter, platforms=zhipin) but that is a precondition, not usage guidance. There is no statement of when to pick this over boss_agent_run, boss_agent_review, or boss_agent_pending, and no mention of what happens after training completes.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are supplied, so the description carries the entire behavioral burden and only partially discharges it: it discloses the output shape (score plus gap analysis). It says nothing about cost, latency, quota consumption, whether the call is read-only, or how resume names are resolved, all of which matter for an AI-backed operation with no readOnlyHint.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

The core action and its output are front-loaded in a single compact clause with no filler. The trailing bracketed availability block is dense but earns its space as the only access-control signal present.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

With no output schema and no annotations, the description at least names the return contents (match score and gap analysis), which covers the main gap. It stops short of describing result granularity, output language/format, or failure behavior when the named local resume cannot be found.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, with both parameters already documented ('职位描述文本' and '对比的本地简历名称'), so baseline 3 applies. The description adds no further meaning beyond the schema — it does not clarify expected length of jd_text or how the resume name is matched against boss_resume_list output.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific compound action: analyze a job description, score resume match, and output gap analysis, which tells an agent exactly what comes back. However, it never distinguishes itself from close siblings such as boss_ai_fit, boss_ai_optimize, or boss_ai_suggest, so an agent must guess which AI-analysis entry point to pick.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability tag (roles=candidate; candidate_platforms=zhipin; recruiter_platforms=-) effectively says who must not use it, which is real exclusion guidance. But it gives no positive conditions — no mention of when to prefer this over boss_ai_fit or boss_ai_optimize — so usage is only implied.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations exist, so the description must carry the behavioral burden. It does disclose what the tool produces (a diagnosis, action suggestions, and ready-to-send templates), which tells the agent this is a generative advisory call rather than a lookup. It says nothing about whether it mutates state, how long it takes, or whether the chat text is sent to a model, leaving the side-effect profile to inference.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

It is a single front-loaded clause stating the function, followed by a compact availability constraint. No sentence is wasted. The availability metadata is slightly cramped inside brackets but remains readable.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a three-parameter generative tool with no annotations and no output schema, the description covers the purpose and the return shape at a high level. It omits the one thing an agent most needs in this crowded AI family: how this differs from boss_ai_reply and the other advisory siblings.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so style, resume, and chat_text are already documented with examples in the schema. The description adds no further meaning about input format, expected length of chat_text, or how style influences output. With the schema doing the heavy lifting, the baseline of 3 is appropriate.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description names a concrete verb-resource pair: it diagnoses communication state from chat logs and returns next-step actions plus ready-made message templates. That is specific and actionable. It stops short of differentiating itself from AI siblings such as boss_ai_reply or boss_chat_summary, so an agent cannot fully rule those out from the text alone.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability block is genuinely useful: it states this only applies to the candidate role on zhipin and is explicitly unavailable for recruiters, which prevents a whole class of wrong calls. It gives no guidance, however, on when to prefer this over boss_ai_reply or boss_chat_summary, so the choice among AI chat tools remains implied rather than stated.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries full burden; it does disclose the key behavioral trait that output is draft-only and never sent, which prevents false expectations of an apply action. It omits what the draft contains, length, formatting, and failure behavior when job_id/jd_text is missing.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

A single front-loaded sentence pairs the action with its inputs and the draft-only constraint; the bracketed availability tag is compact. Minimal waste overall.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a moderate-complexity generative drafting tool with no output schema, the description covers purpose, inputs, and the critical not-sent constraint. Missing only draft-format expectations and error/enumeration behavior, which are minor.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100% with per-field Chinese descriptions and enums for lang and tone, so the schema already carries parameter meaning. The description only echoes the resume/job inputs and adds no syntax or precedence detail (e.g. job_id vs jd_text exclusivity).

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb (起草) and resource (求职信/自我介绍), plus scope (based on local resume and target job). It clearly distinguishes itself as a drafting tool, though it does not name which sibling handles adjacent tasks like optimization or suggestions.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The availability tag (roles=candidate, candidate_platforms=zhipin) gives real routing context, and 'based on local resume and target job' implies the input precondition. However, no explicit alternatives are named (e.g. boss_ai_suggest, boss_ai_optimize) and there is no when-not guidance.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are supplied, so the description must carry the full behavioral burden. It helpfully discloses the preconditions (a locally stored résumé and already-cached pool job details) and the customer-facing platform restriction, but says nothing about whether the analysis is read-only or persists a report, whether it consumes AI quota/latency, or how failures surface when cached details are missing.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

One front-loaded sentence states the data sources and the outputs, followed by a compact availability tag. There is no padding, though the availability tag is somewhat telegraphic and could be clearer to an agent unfamiliar with the roles/platform vocabulary.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

With only two parameters and no output schema, the description covers the essentials: inputs, generated artifacts, and platform availability. It still leaves the agent guessing about the report's shape and whether results are returned inline or persisted, which matters for a tool whose whole purpose is producing a report.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so both parameters are already documented (resume = 简历名称, limit = max pool jobs analyzed). The description adds mild context by stating the analysis is driven by a local résumé across cached pool jobs, which supports the meaning of `resume`, but it adds no detail on the `limit` default behavior or résumé-name resolution beyond the schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description names a specific verb (生成/generate) and a concrete output set — per-job match degree, capability gaps, and keyword-hit report — together with its data sources (local resume plus cached job details in the candidate pool). This is clearly separable from analytical siblings such as boss_ai_analyze_jd or boss_ai_resume_optimize, though no sibling is named explicitly.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability constraint (roles=candidate, candidate_platforms=zhipin, recruiter_platforms=-) is genuinely useful routing information that tells an agent when this tool is callable. Beyond that, it only implies usage — pool jobs must already be cached — and never states alternatives or exclusions, e.g. when to choose this over boss_ai_analyze_jd or boss_shortlist_compare.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are supplied, so the description carries the burden. It usefully discloses the role/platform availability constraint, which is genuine behavioral context. But it says nothing about whether generation is safe/non-mutating, latency, rate limits, or what the output looks like.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

Single front-loaded sentence plus a bracketed availability tag; no wasted wording. The bracketed constraint is compact though somewhat cryptic in format.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a generation tool with no output schema and no annotations, the description covers purpose and availability but leaves the return shape (number/format of questions, advice) and any safety profile unstated. Adequate but with clear gaps.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so the baseline is 3. The description does add that supplying a resume customizes the generated questions, which clarifies the purpose of the optional resume param beyond the schema's '参考简历名称'. Marginal added value.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource: generates mock interview questions and preparation advice from a target JD, with optional resume-based customization. Clear on its own, but it doesn't differentiate from adjacent AI siblings like boss_ai_fit or boss_ai_chat_coach.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The availability line (roles=candidate; candidate_platforms=zhipin) gives real gating context, and 'based on target job description' implies when it applies. However, there is no explicit when-not guidance or naming of alternative tools such as boss_ai_fit for prepping differently.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description still carries useful behavior: the operation is non-destructive because it returns an optimized structure rather than writing to disk. That is the single most valuable disclosure here. It does not say whether the optimization is persisted anywhere, whether it costs credits/is rate-limited, or what the output object contains beyond 'structure'.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

One front-loaded sentence plus a compact availability tag; nothing is padded. The availability metadata is dense but readable and communicates real preconditions rather than boilerplate.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a two-parameter, no-annotation, no-output-schema mutation-like tool, the definition covers intent and the no-write guarantee, which is the critical part. It leaves the return shape vague ('optimized structure') and never explains how to obtain a valid resume identifier, so an agent could stall on inputs.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so both parameters are documented in the schema and the baseline is 3. The description adds that jd_text is the target job description being optimized against, which gives purpose but no format or length guidance, and it does not clarify where the 'resume' name value comes from (boss_resume_list/boss_resume_show).

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description states a specific verb and resource ('optimize resume based on target job description') plus a scoping clarification ('outputs optimized structure, does not write to disk'). However it does not distinguish itself from the near-identical sibling boss_ai_resume_optimize, so an agent still has to guess which of the two to pick.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The availability bracket gives real gating (roles=candidate, candidate_platforms=zhipin, recruiter_platforms=-), which tells the agent when the tool is usable. It says nothing about when to prefer it over boss_ai_resume_optimize, boss_ai_analyze_jd, or boss_ai_fit, and provides no prerequisites such as needing a resume name from boss_resume_list.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the full behavioral burden. It usefully discloses that 2-3 candidate drafts are produced, not a single message, and that this only works for the candidate role on the zhipin platform. It does not state whether drafts are ever sent automatically, whether there are rate limits, or what the drafts contain, so important behavioral gaps remain.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

The definition is a single compact sentence front-loaded with the action and the output shape, followed by a clearly bracketed availability constraint. No sentence is wasted, though the bracketed metadata is opaque to a reader unfamiliar with the shorthand.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

The essentials are covered: purpose, output cardinality, and platform/role gating for a 4-parameter tool with a fully documented schema. Because there are no annotations and no output schema, a stronger definition would spell out the draft format and confirm the tool has no side effects; the definition stops short of that.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so all four parameters, including the tone enum and the optional resume and context fields, are already documented in the schema. The phrase 支持简历参考和语气偏好 only restates the resume and tone parameters and adds no new meaning such as expected formats or how many resume names are accepted.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description gives a specific verb and resource: generating reply drafts (回复草稿) from a recruiter message, and it states that 2-3 candidate drafts are produced. The availability tag (roles=candidate, recruiter_platforms=-) implicitly separates it from the recruiter-side sibling boss_hr_reply, but this separation is not stated in prose.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The intended usage is implied: generate a reply when a recruiter has messaged you, with resume reference and a tone preference. However, no explicit when-to-use, when-not-to-use, or alternative guidance is given, such as a pointer to boss_ai_chat_coach for live coaching or to boss_hr_reply for the recruiter side.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are supplied, so the description carries the full behavioral burden. It does disclose two meaningful traits: the operation is suggestion-only and does not modify the resume, and it is gated to candidate role on the zhipin platform. It says nothing about output format, how the resume content is obtained, or any cost/rate limits.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single compact sentence plus an availability tag, with the core purpose front-loaded. Nothing is wasted, though the bracketed availability metadata is somewhat cryptic in placement.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 3-parameter advisory AI tool with no output schema and no annotations, the description covers intent and the non-mutating nature but leaves return-value expectations and resume-source behavior unstated. It is minimally sufficient but not thorough.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so job_id, jd_text, and their mutual exclusivity are already documented in the schema. The description adds no semantics about parameters beyond what the structured fields already state, so the baseline of 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description names a specific verb and resource: it optimizes resume wording based on the target position, and it clarifies the operation is advisory only ('仅建议，不修改简历'). It does not, however, explicitly distinguish itself from similarly named siblings such as boss_ai_optimize or boss_ai_suggest, so an agent must infer the difference from the scope note.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

There is no explicit when-to-use/when-not-to-use guidance or mention of alternatives. The embedded availability constraint (roles=candidate; candidate_platforms=zhipin; recruiter_platforms=-) does imply who may call it, which is useful, but eligibility is not the same as usage guidance.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations supplied, the description carries the full burden and does disclose the two traits that matter most: the operation is non-mutating and the output is priority-ordered suggestions rather than an edited resume. Role/platform gating is also surfaced. It stops short of describing rate limits, cost, or what the suggestion payload looks like.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

One compact sentence plus a bracketed availability tag, with the core purpose and the non-mutation guarantee front-loaded. The availability metadata is dense but earns its place as the only usage constraint given.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a two-parameter, read-only advisory tool with full schema coverage and no output schema, the description covers purpose, result ordering, non-mutation, and audience gating. Only the shape of the returned suggestions is left unstated, which is a minor gap for a tool this simple.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, with both 'resume' and 'jd_text' documented in the schema itself, so the baseline is 3. The description's '基于目标职位' only confirms that jd_text drives the analysis and adds no format or linkage detail beyond that.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource ('给出简历改进建议' – produce resume improvement suggestions) scoped to a target position, and explicitly negates mutation ('不修改简历'), which cleanly separates it from the sibling optimize tools (boss_ai_optimize, boss_ai_resume_optimize). An agent can route on this without opening any schema.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The availability tag ('roles=candidate; candidate_platforms=zhipin; recruiter_platforms=-') gives explicit eligibility conditions, and the 'does not modify resume' clause signals when to prefer this over an optimizing sibling. It does not, however, name those siblings directly or state prerequisites such as needing a saved resume or a JD first.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations exist, so the description carries the full burden. It does disclose meaningful preconditions – that the input is drawn from the candidate pool and that the tool is only available in the candidate role on the zhipin platform – but it never states that the operation is read-only, whether it consumes quota or triggers model calls, or what shape the recommendations take.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

A single front-loaded purpose clause followed by a compact bracketed constraint block – no wasted wording. The bracketed availability syntax is terse but decodable.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a one-parameter AI-suggestion tool with no output schema, the definition covers purpose and platform/role gating but leaves the return value, invocation cost, and downstream use of the keywords unexplained. Adequate to call correctly, not sufficient to use confidently.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

There is only one parameter and schema coverage is 100%, with 'limit' fully described as the cap on candidate-pool jobs analyzed. The description adds no meaning beyond that, so the baseline of 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description states a specific verb and output ('推荐搜索关键词组合' – recommend search keyword combinations) plus its basis ('基于候选池职位分析' – analysis of jobs in the candidate pool). This distinguishes it from generic siblings like boss_ai_suggest or boss_ai_optimize, though it never names an alternative explicitly.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability line gives a real gating condition (roles=candidate, candidate_platforms=zhipin, recruiter_platforms=-), telling the agent when the tool is even callable. However, there is no guidance on when to prefer this over boss_ai_suggest or how to act on the returned keywords.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the full behavioral burden, and it does disclose one genuinely useful trait: the action is idempotent and will not re-submit a duplicate application. It says nothing about side effects (does it also open a chat?), error conditions, or whether a message is sent to the recruiter.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

A single compact sentence with the action front-loaded, followed by a clearly delimited availability tag. No filler, though the bracket syntax is machine-ish rather than explanatory.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a two-parameter mutation tool with no annotations and no output schema, the description covers what the action is, its idempotency, and its eligibility constraints. It is still incomplete about the outcome of the call and the relationship to the greeting tools, which matters given how many siblings overlap.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100% and both parameters (job_id as encrypt_job_id, security_id) are documented in the schema itself. The description adds no further meaning about how the two IDs relate or where they come from, so baseline 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description states a concrete action pair, "发起投递或立即沟通" (apply to a job or start an immediate conversation), which tells an agent exactly what operation this is. The only weakness is that "立即沟通" overlaps conceptually with the sibling boss_greet, and the description never disambiguates the two.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability tag (roles=candidate, candidate_platforms=zhipin, recruiter_platforms=-) gives a real precondition for when the tool can be used at all. However, there is no guidance on when to choose this over boss_greet or boss_batch_greet, so the agent is left to infer the boundary.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the full burden, and it does disclose the batch cap of 10, which is a real behavioral constraint. It does not disclose that greeting sends real outbound messages to real users, whether the operation is idempotent (re-greeting someone already contacted), or any rate limits — significant omissions for an irreversible outbound action.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

A single front-loaded clause plus a bracketed availability tag — no filler. The cap is placed prominently rather than buried, though the mixed-language bracketed suffix is slightly dense.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a batch, side-effecting outbound-messaging tool with four parameters, no annotations, and no output schema, the description is thin. It never explains dry_run (the agent must infer 'preview mode' from the schema), what the response looks like, or how partial failures are handled — all things the agent needs before firing a batch of greetings.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the baseline is 3. The description does add one piece of meaning beyond the schema by stating the upper bound of 10, which the schema's '最大打招呼数量' field never specifies. It adds nothing about query, city, or dry_run, so it does not rise above the baseline.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and scope: batch greeting after a search, with a hard cap of 10. The 'batch' qualifier implicitly distinguishes it from the single-greet sibling boss_greet and the recruiter-side boss_hr_greet, though neither sibling is named explicitly.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The availability tag (roles=candidate; candidate_platforms=zhipin; recruiter_platforms=-) is a genuine gating condition that tells the agent when the tool is usable. However, there is no guidance on when to prefer this over boss_greet for a single target, or on what to do when the search returns fewer than the requested limit.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the full burden. It discloses useful traits: this is a read/list operation, results carry a stable identifier, and availability is constrained to candidate role on the zhipin platform with no recruiter platform support. It does not disclose ordering, pagination behavior beyond the 'page' param, or whether results are truncated, which matters for a data-listing tool with no output schema.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

Two compact clauses plus one bracketed availability note — every element carries information. The core capability is front-loaded and the cross-tool uid handoff follows immediately, so an agent gets the essential facts in the first read.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a three-parameter, all-optional list tool with no output schema, the description covers the key gaps: what is listed, which filters exist, one critical return field (uid), and its use in sibling tools. The main shortfall is the absence of any hint about what else each item contains, which an agent might need before deciding to enumerate conversations.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so the schema already documents 'days', 'page', and the 'from_who' enum. The description only restates the filtering capability at a high level ('按发起方和时间筛选') without adding format, default, or interaction details beyond what the schema provides. Baseline 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description states a specific verb and resource ('查看沟通列表' – view the chat/conversation list) and names the supported filters (initiator and time). It also distinguishes itself from message-level siblings by explaining that the returned 'uid' feeds into boss_chatmsg / boss_mark / boss_exchange, which tells an agent this is the list-level entry point rather than a content reader.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

It implies when to use it (listing conversations, optionally narrowed by sender or recency) and gives a concrete downstream workflow via uid. However, it never states when NOT to use it or names explicit alternatives — e.g. boss_chat_summary for aggregated views or boss_chatmsg for reading the actual messages — and the role restriction is buried in a bracket rather than framed as a usage condition.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are supplied, so the description bears full behavioral burden. It discloses role/platform availability, which is genuinely useful, but says nothing about permissions, pagination behavior, rate limits, or the compliance gating referenced only in the schema's 'raw' parameter. For a data-read tool with zero annotation coverage, more disclosure would be expected.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

A single sentence plus a compact availability tag, with purpose front-loaded before constraints. Nothing is wasted and nothing redundant repeats the title or schema.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

The description covers purpose and eligibility for a four-parameter read tool with full schema coverage and no output schema. It omits what the response looks like (and how 'raw' changes it) and any note on pagination defaults, leaving an agent to infer message-retrieval behavior from the schema alone.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so all four parameters (raw, page, count, security_id) are already documented in the schema, including the uid-vs-security_id guidance. The description adds no parameter-level meaning beyond that, which is acceptable at baseline.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource: '查看与指定好友的聊天消息历史' (view chat message history with a specified contact). An agent can tell this retrieves messages rather than resumes or chat lists. It doesn't explicitly distinguish itself from the sibling boss_hr_chatmsg, though the roles=candidate tag implicitly separates them.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability block gives concrete eligibility constraints: roles=candidate, candidate_platforms=zhipin, recruiter_platforms=-, which tells the agent which callers should and should not use it. It stops short of naming an explicit alternative for recruiters (boss_hr_chatmsg), so routing requires inference.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are supplied, so the description carries the full burden. It discloses the input source (chat history) and what is produced (a structured summary and next-step suggestions), but says nothing about whether an AI model is invoked, whether calls are slow or costly, whether the operation is read-only, or what the output format looks like.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

One sentence plus a bracketed availability tag: the purpose comes first and nothing is wasted. The bracketed metadata is terse to the point of being cryptic, but it is compact rather than verbose.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

No output schema exists, so the description should have described the shape of the summary and suggestions, yet it only names them. Combined with the absence of annotations on a model-backed, possibly slow operation, the definition is adequate but leaves real gaps.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

There is only one parameter, and the schema documentation covers 100% of the input schema; it even explains the recommended uid source from boss_chat and cross-request stability. The description adds no meaning beyond the schema, so the baseline of 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

Names a specific verb and resource: generate a structured summary plus next-step suggestions from chat history. The availability clause (roles=candidate, candidate_platforms=zhipin, recruiter_platforms=-) further narrows what this tool is for. It does not name a sibling to contrast with, even though boss_ai_chat_coach and boss_chat are near neighbors, so it falls short of a 5.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The availability tag gives a real usage gate: it applies only to candidate-role users on zhipin, so recruiters must not call it. Beyond that it offers no explicit when-to-use or when-not-to-use guidance and never points to an alternative such as boss_ai_chat_coach for coaching or boss_chatmsg for raw messages.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the full burden, and it does disclose the access constraints (roles and platform variants) plus the approximate list size and the '全国' sentinel. It does not say whether the list is static or live, whether it depends on the caller's config, or what the entries look like.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

A single compact line: purpose first, then count and special case, then the availability constraint in a bracketed suffix. Dense but front-loaded, with no filler.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a trivial zero-parameter list tool this is close to sufficient, but with no output schema the description should say what the returned list contains (city names only, or name/id pairs) and that those values feed into search filters. As written, an agent knows what the tool is but not exactly what it gets back.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The tool takes zero parameters, so there is nothing for the description to disambiguate; baseline 4 applies. The mention of the '全国' pseudo-city is a useful hint about the value domain even without parameters.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource: '列出支持的城市列表' (list supported cities), and adds concrete scope detail — roughly 44 entries including the special '全国' (nationwide) option. No sibling tool offers a city list, so no disambiguation is needed, but it stops short of saying what the entries are used for.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability tag tells the agent which roles and platforms may call it (candidate/recruiter; zhipin / zhipin-recruiter), which is real usage gating. However, it never says when to call this versus other tools — e.g. that it should be used to discover valid city values before calling boss_search or boss_recommend.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are supplied, so the description carries the full behavioral burden and does very little with it. It never states that files are permanently deleted, what exactly is considered 'expired', whether the operation is reversible, or whether it can affect active sessions/data; the only safety affordance (dry_run) is visible solely through the schema.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

A single compact sentence with the availability gate appended; nothing is padded and the action is front-loaded. The bracketed key=value notation is terse but readable.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a destructive maintenance tool with no annotations, no output schema, and no indication of what gets removed or what the result looks like, the description is thin. An agent could invoke it but could not predict or explain the consequences of doing so.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Both parameters (all, dry_run) are fully documented in the schema with defaults, so the schema does the heavy lifting and the description adds no further meaning. Baseline 3 applies given 100% schema description coverage.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description gives a specific verb and resource – cleaning expired cache and temporary files – which is unambiguous and distinguishable from every sibling tool (none of which is a maintenance/cleanup command). It lacks any statement of scope or effect, but the core action is clear.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed block states availability constraints (roles=candidate/recruiter, platform=zhipin/zhipin-recruiter), which is a precondition rather than a when-to-use rule. There is no guidance on when cleanup is warranted, how it differs from boss_doctor or boss_status, or any warning about running it.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are supplied, so the description carries the full behavioral burden. It does not disclose that 'set' and 'reset' mutate configuration, whether 'reset' restores defaults or deletes the key, whether changes are reversible, or what permissions are required beyond role eligibility. Only the availability envelope is covered.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

A single short sentence plus a bracketed constraint block, front-loaded with the purpose and free of filler. The densest useful information (actions and their effects) is absent rather than the text being bloated, so conciseness itself is good.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

This is a low-parameter but mutation-capable tool with no annotations and no output schema, so the description should explain what 'reset' does, what a 'set' returns, and whether keys are namespaced. None of that is present, leaving an agent unable to predict the consequences of the destructive actions.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100% — every parameter including the action enum has its own Chinese description — so the baseline is 3. The description adds nothing about parameter semantics beyond what the schema already states, such as whether 'key' must exist before 'get' or what 'reset' falls back to.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description states a specific verb pair and resource ('查看和修改配置项' = view and modify configuration items), which tells an agent what domain the tool covers. It does not distinguish this tool from adjacent configuration-like siblings such as boss_preset_list or boss_watch_list, so it stops short of full sibling differentiation.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability block gives concrete eligibility constraints (roles=candidate/recruiter; platform=zhipin/zhipin-recruiter), which is useful routing information. However, it never says when to pick this tool over the preset/watch siblings, and it does not explain which action (list/get/set/reset) is appropriate in which situation.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are supplied, so the description carries the full disclosure burden. It does add useful context that results are persisted and can be split by detail-fetch state (completed vs. pending), and it flags platform/role gating. It stops short of describing pagination behavior, ordering, or what happens with an unknown/expired run_id.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

Two compact clauses, with the core action front-loaded before the availability qualifier. No filler text, though the bracketed availability string is dense and would read better as a plain sentence.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a simple read-only retrieval tool with three fully described parameters and no output schema, the description covers purpose, filtering, and applicability. The main gap is relational context — how run_id ties back to boss_crawl_start/boss_crawl_status — which the agent must infer.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so all three parameters, including the enum values, are already fully documented in the schema. The description only restates the filter dimensions and adds no syntax, format, or edge-case meaning beyond that, making 3 the appropriate baseline.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource: read persisted crawl job results, with page and detail-status filtering. An agent can tell this returns stored crawl output rather than live searches, though the description never distinguishes it from nearby siblings such as boss_crawl_status or boss_crawl_shortlist.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability block gives real scoping guidance (roles=candidate, candidate_platforms=zhipin, recruiter_platforms=-), which tells an agent when this tool is applicable. However, there is no explicit when-to-use vs. alternative guidance (e.g., use boss_crawl_status for run state, this for the resulting jobs), so the choice among crawl siblings is left implicit.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations supplied, the description carries the full burden; it usefully discloses that no BOSS request is made, meaning this is a local-only write. It does not disclose whether existing shortlist entries are overwritten, whether duplicates are skipped, what happens to stale selectors, or what the call returns after import.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

The core action and the no-network trait are front-loaded in one sentence, which is efficient. The appended availability bracket is terse and machine-formatted but earns its place as a hard eligibility gate.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a five-parameter mutation tool with no annotations and no output schema, the description covers purpose and local-only behavior but leaves the post-import result undefined. An agent still cannot tell whether the call reports how many positions were added, skipped, or rejected.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so all five parameters, including the all/selectors mutual exclusion and the run_id provenance, are already documented in the schema. The description only restates that selectors originate from crawl results and adds no format, sizing, or precedence guidance beyond the schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description names a specific verb and resource ('将 crawl results 返回的 selector 导入本地 shortlist') and adds a distinguishing trait ('不请求 BOSS'), so the agent knows this is a purely local import from crawl output rather than a fresh fetch. It does not explicitly contrast itself with nearby siblings such as boss_shortlist_add or boss_crawl_results, leaving the boundary to be inferred.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

Usage is implied through the referenced workflow (selectors come from boss_crawl_results, run_id from boss crawl start), and the availability bracket states a when-not condition for recruiter roles. However, there is no explicit statement of when to prefer this over boss_shortlist_add or what to do when a run has no importable positions, so the agent must infer the choice.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are supplied, so the description carries the full burden. It does usefully reveal the return surface (cursor, job count, progress, risk status), which is real behavioral context given there is no output schema, and it implies a read-only monitoring operation. It does not describe polling cadence, how progress or risk status are encoded, or what happens for an unknown or finished run_id.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

A single front-loaded sentence listing exactly what is returned, followed by a bracketed availability tag. Every element earns its place and there is no filler or redundancy.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a one-parameter read tool with no output schema, listing the returned fields plus the role/platform availability gate covers what an agent needs to call it and interpret the result. Minor gaps remain around polling behavior and the meaning of risk status values, but nothing blocking.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Only one parameter and schema coverage is 100% – the schema already defines run_id as the crawl-run identifier returned by boss crawl start, so this is inherently well covered. The description adds nothing beyond that, but with full schema documentation the parameter is unambiguous as-is.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb (读取/read) plus the concrete resource fields it exposes: crawl 页游标 (page cursor), 职位数 (job count), 详情进度 (detail progress) and 风险状态 (risk status). That is enough to separate it from result-oriented siblings like boss_crawl_results and boss_crawl_shortlist, but the description never names those siblings to make the distinction explicit.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

Usage is only implied: the run_id schema note ('由 boss crawl start 返回') tells the agent this is a follow-up call after starting a crawl, and the availability tag restricts it to candidate role on the zhipin platform. There is no explicit when-to-use vs when-not statement, and no guidance on calling it repeatedly to poll progress versus using boss_crawl_results to fetch output.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are supplied, so the description carries the full burden. It discloses the availability constraints (candidate role, zhipin platform only) and implies a read-only lookup, but says nothing about permissions, rate limits, error cases when security_id is stale, or what the response contains. Useful but incomplete behavioral context for an unannotated tool.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

Two short sentences with the purpose front-loaded, followed by a compact bracketed availability tag. Nothing is redundant, though the bracketed key=value notation is terse to the point of being cryptic for an agent unfamiliar with the schema.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a simple two-parameter read tool with no output schema, the description covers purpose, prerequisite source of the ID, and caller eligibility. The main remaining gap is that it never indicates what the returned job details include, which matters slightly since no output schema exists.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the schema already documents both job_id (fast path via encrypt_job_id) and security_id. The description only adds the provenance of security_id (from search/recommend results) and omits the job_id fast-path entirely. Baseline 3 is appropriate when the schema does most of the work.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

"查看职位详情" gives a specific verb (查看) and resource (职位详情), so the agent can tell it reads job posting details. The availability tag (roles=candidate; recruiter_platforms=-) implicitly separates it from the recruiter-side sibling boss_hr_jobs_detail, though the description never names that sibling directly.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

It states where the required argument comes from ("从 search/recommend 结果获取"), which tells the agent to run boss_search or boss_recommend first, and the availability tag scopes who may call it. It does not explicitly say when not to use it or how it differs from boss_hr_jobs_detail beyond the platform field.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations supplied, the description must carry the full burden. It does disclose the important behavioral traits: the operation is read-only (只读), it draws from three data domains, and md output is intended for direct email/Feishu delivery. It does not say whether results are scoped to the calling candidate account, how large the digest is, or whether empty sections are omitted.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

One dense sentence with the payload stated first, followed by the output-format rationale and a compact bracketed availability tag. No filler or repetition, though the availability bracket is dense enough to slow parsing slightly.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

All three parameters are optional with 100% schema coverage and there is no output schema, so the description only needs to frame what comes back — which it does at section level. It stops short of describing the digest's shape (counts vs. lists), so an agent cannot predict response structure or size before calling.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so the baseline is 3. The description adds modest value by explaining why format=md matters (邮件/飞书直发, i.e. paste-ready delivery), which contextualizes the enum beyond the schema's '输出格式'. It adds nothing about output or days_stale that the schema's own descriptions do not already say.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description names a specific verb (汇总 / aggregate) and a specific artifact (只读日报 — a read-only daily digest) and enumerates its three constituent sections: new jobs, follow-up-pending conversations, and interview items. This clearly distinguishes it from single-purpose siblings like boss_follow_up, boss_interviews, and boss_stats, even though those siblings are not named explicitly.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

Usage is implied rather than stated: an agent can infer this is the combined daily roll-up, and the bracketed availability gate (roles=candidate; candidate_platforms=zhipin; recruiters unsupported) is genuinely useful eligibility information. However, there is no explicit 'use X instead when Y' routing against boss_follow_up, boss_stats, or boss_chat_summary, and no mention of when a digest is inappropriate.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the full behavioral burden. The enumerated check areas (environment, dependencies, login state, connectivity) give real insight into what the tool inspects, but it never states that the operation is read-only/non-mutating, whether it hits the network, whether it is safe to run repeatedly, or what the result looks like.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

One front-loaded sentence with no filler, plus a compact machine-readable availability tag. The bracketed roles/platforms metadata is dense but earns its place by gating access. Slightly terse to the point of omitting any output or safety statement.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a zero-parameter diagnostic tool with no annotations and no output schema, the description should at minimum characterize the result an agent will get back, since the return value is the entire product of a diagnostic. It lists what is checked but says nothing about result shape, severity levels, or remediation hints, leaving a meaningful gap.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The tool takes zero parameters and the schema is an empty object, so there is nothing for the description to disambiguate. Baseline 4 applies for a parameterless tool.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description gives a specific verb (诊断/diagnose) and enumerates the concrete resources checked: local runtime environment, dependencies, login state, and network connectivity. This is far more informative than the bare name 'boss_doctor'. It does not, however, distinguish itself from the adjacent sibling boss_status, which an agent could reasonably confuse with it.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

Usage is only implied: a diagnostic tool is for troubleshooting when something is broken, and the bracketed availability clause tells the agent which roles and platforms it applies to. There is no explicit statement of when to call it (e.g. before a first run, or after a failed login) nor any exclusion pointing to boss_status or boss_config as the alternative.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

There are no annotations, so the description carries the full disclosure burden. Using '请求' (request) rather than a get-style verb conveys that this initiates an exchange request rather than revealing contact data immediately, which is meaningful. It still omits side effects, whether the counterparty must approve, quota or rate limits, and what the call returns.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

One sentence plus a compact constraint tag, with the action front-loaded and zero redundant prose. The bracketed availability notation is terse but functional. Its machine-oriented style slightly reduces readability for an agent scanning natural language.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a single-parameter request tool with no annotations and no output schema, the definition covers the action and platform gating but not the outcome. An agent cannot tell whether the call returns contact details, queues a request pending approval, or returns only a status. The missing output schema should have been compensated for and was not.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100% and there is only one parameter (security_id), whose schema text already explains that it can be the contact's uid or security_id. The description adds no constraint or format guidance beyond that, so the baseline 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description gives a specific verb-and-resource pair: '请求交换联系方式（手机号或微信）' (request exchange of phone number or WeChat). This is clearly a distinct action from resume viewing or greeting. It does not explicitly name a sibling, but the availability line marks it as candidate-only, which implicitly separates it from the recruiter-side boss_hr_exchange.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability block specifies the usable context (roles=candidate, candidate_platforms=zhipin, recruiter_platforms=-), which is real when-to-use guidance. However, it never says when NOT to use this or which alternative to pick—for example, the recruiter counterpart boss_hr_exchange—leaving the agent to infer routing from the fact that recruiters are excluded.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the full burden and delivers meaningful behavior: default redaction of job_id/security_id/boss_name, unconditional omission of platform marker, recruiter name and salary in HTML, and different return channels depending on whether output_file is supplied. It does not mention permissions beyond the role gate, rate limits, or whether a fetch/crawl is triggered as a side effect.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

Front-loads the core action and output formats, then layers redaction and return-shape details in a compact sequence. Dense but every clause (formats, URL reuse, redaction, envelope behavior) carries information; the bracketed availability tag is slightly clunky but useful.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 14-parameter tool with no output schema and no annotations, the description covers the essentials an agent needs: output formats, file-vs-envelope return behavior, and privacy defaults. Missing pieces are secondary (result caps, error behavior, auth beyond the role tag).

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the schema already documents all 14 parameters and the baseline is 3. The description adds real meaning for a few (url supplanting query, include_private scope, output_file changing the return envelope) but leaves the many filter parameters to the schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

Opens with a specific verb-plus-resource (导出搜索结果为 CSV / JSON / HTML 文件) and names the data source (BOSS 直聘搜索页), which clearly separates it from siblings like boss_search or boss_crawl_results. An agent can tell what it produces and in what formats without opening the schema.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

Establishes the usage context (exporting existing search results, reusing a search-page URL to carry filters) and states the availability constraint roles=candidate / platform=zhipin. It stops short of explicitly naming alternatives or saying when not to use it versus boss_search or boss_crawl_results.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description still discloses the key behavior '不落库' (does not persist to the database), which tells the agent this is a side-effect-free read. It also indicates that '有效状态' (validity status) is included in the result. It stops short of describing pagination behavior, total page counts, or what happens at out-of-range pages.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

One compact sentence with the core purpose front-loaded, followed by a bracketed constraint tag. Nothing is redundant, though the dense bracket notation is less scannable than prose.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a low-complexity read tool with one optional parameter and no output schema, the description covers purpose, non-persistence, returned status, and platform eligibility. Only the paging mechanics are under-specified, which is a minor gap.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

There is a single optional integer parameter 'page' with 100% schema description coverage, so the schema already carries the semantics. The description's phrase '单页' reinforces that results are paged but adds no default value, range, or 1-based/0-based convention beyond what the schema states.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description names a specific verb ('预览'/preview) and a specific resource ('BOSS 职位收藏' job favorites list) plus what is returned ('单页及有效状态'). It is clearly distinguishable from siblings such as boss_shortlist_list or boss_watch_list by the 'favorites' domain, though it never names those alternatives explicitly.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability block ('roles=candidate; candidate_platforms=zhipin; recruiter_platforms=-') does give a precondition for when this tool is callable, which is useful gating. However, there is no guidance on when to use this versus the other list-style siblings, and no note on how to page through results.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are provided, so the description carries the full disclosure burden. It implies a read-only filtered listing and defines the three qualifying conditions, but says nothing about ordering, result volume, pagination, or whether stale-day logic interacts with interview criteria. That leaves real behavioral gaps for a no-annotation tool.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

One compact sentence states the purpose and the criteria, followed by a bracketed availability tag. Nothing is redundant and the core function is front-loaded.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

There is no output schema, so the description should say something about what a candidate entry contains or how many are returned; it does not. The input side is complete thanks to full schema coverage, but the return-side picture is left entirely to inference.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, with days_stale fully documented including its default of 3, so the schema does the heavy lifting. The description's mention of '超时未推进' hints at the staleness concept but adds no syntax, unit, or interaction detail beyond the schema. Baseline 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description uses a specific verb ('筛出' / filter out) and names the resource explicitly ('候选项' / candidates) plus the three inclusion criteria (unread, stale, interview). An agent can grasp the function without opening the schema. It stops short of differentiating from nearby siblings such as boss_pipeline or boss_hr_last_messages, which could also surface follow-up candidates.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability line gives concrete scoping: roles=candidate, candidate_platforms=zhipin, recruiter_platforms=-, which tells the agent both who this is for and where it does not apply. It does not, however, name an alternative tool or state an explicit when-not condition beyond the platform/role gating.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are supplied, so the description carries the full disclosure burden for what is clearly an outbound action that sends a message to a third party. It never states whether the greeting consumes a daily quota, whether it is idempotent, what happens if the candidate has already greeted that job, or what error states look like. The only behavioral hint is the availability gate, which is insufficient for a mutating outreach tool.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

Two short sentences plus an availability tag, with the action front-loaded. The middle sentence restates the schema's required list verbatim, which is mild redundancy but not enough to harm readability.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a two-parameter action tool with no output schema, the description covers the action, the required inputs, and platform/role availability. It is still missing side-effect and failure-mode information that an agent needs before triggering outbound contact, and it does not point to sibling alternatives.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%: job_id is documented as the encrypt_job_id and security_id as the position's security_id, so the schema does the heavy lifting. The description only restates that both are required, adding no format, sourcing, or derivation guidance (e.g., where security_id comes from). Baseline 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

States a concrete verb+resource: greeting a recruiter ('向招聘者打招呼'), which is immediately distinguishable from generic chat or search tools. It does not, however, distinguish itself from close siblings such as boss_hr_greet (recruiter-side) or boss_batch_greet (batch variant), leaving the agent to infer the difference.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability block gives a real usage gate: only usable when roles=candidate, only on candidate_platforms=zhipin, and explicitly unavailable to recruiter platforms. That is a genuine when-to-use condition. It stops short of naming alternatives (e.g., boss_batch_greet for multiple jobs), so no explicit when-not guidance.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations at all, the description carries the full behavioral burden. '查看' implies a read-only operation and the availability clause discloses a role/platform precondition that will cause failures for recruiters, which is genuinely useful. However, nothing is said about pagination, result volume, ordering, or whether any state is recorded when the history is read — gaps that matter for a tool with zero annotation coverage.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

One short clause plus a bracketed availability tag — front-loaded and free of filler. The bracketed metadata is dense but machine-parseable rather than prose padding, so it earns its place.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a zero-parameter, no-output-schema read tool the description is close to sufficient, and the availability gate is a valuable addition. What is missing is the scope of the returned history — which entity's browsing history, how far back, and whether results are truncated or paginated — which an agent needs in order to interpret the result.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

The tool takes no parameters and the schema is an empty object, so there is no parameter meaning the description needs to supply. Baseline 4 applies; the description neither helps nor harms here.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description gives a specific verb+resource pair: '查看' (view) plus '浏览历史' (browsing history), which is a distinct capability not obviously duplicated by any sibling tool in the list. It stops short of naming what kind of history (e.g. viewed jobs vs. viewed resumes) or differentiating itself from neighbors like boss_watch_list or boss_shortlist_list, so it is clear but not maximally precise.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability block gives real usage conditions — only usable when roles=candidate and candidate_platforms=zhipin, and explicitly unavailable to recruiters (recruiter_platforms=-). That is useful gating information, but there is no guidance on when to prefer this over alternatives or what a caller should do instead when the gate fails.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the full behavioral burden. It usefully discloses that nothing is auto-downloaded and no retry occurs, plus the role and platform constraints, but it says nothing about side effects on the candidate's session, idempotency, or what happens when the request is stale or already handled.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

It is a single dense sentence with the action first, the safety gate second, and availability in a bracketed tag at the end. Nothing is redundant, though the bracket syntax is terse and machine-flavored.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a mutation tool with no annotations and no output schema, the definition covers the approval gate and platform eligibility but leaves out the result of a successful or failed call and the preview semantics of dry_run, which an agent needs before invoking it.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so all four parameters are already documented in the schema. The description adds the approval-sequencing meaning behind yes=true, but says nothing about dry_run or how friend_id and message_id are obtained, so the baseline of 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description states a specific verb and resource: accepting a candidate's attachment-resume request in recruiter mode. It is distinguishable from sibling tools like boss_hr_request_resume and boss_hr_download_resume, but it never explicitly names those alternatives, so the differentiation must be inferred from the word 'accept'.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

It gives a clear gating rule: pass yes=true only after the operator has explicitly approved, and it scopes availability to role=recruiter on zhipin-recruiter. It does not say when to use a different tool, such as requesting the resume rather than accepting it, or how dry_run fits into the workflow.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are provided, so the description carries the full burden. It does disclose a genuine behavioral constraint the schema cannot express — the required recruiter role and zhipin-recruiter platform — which is meaningful context. But it says nothing about result volume, pagination behavior, ordering, or whether viewing applications has side effects (e.g. marking them read), leaving notable gaps for a no-annotation tool.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

Two compact segments: the purpose is front-loaded, followed by a terse availability tag. There is no filler. The bracketed metadata is dense but readable and earns its place by carrying the role/platform gating that no structured field supplies.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

This is a low-complexity paginated list tool with fully documented parameters and no output schema to explain. The description covers what it returns at a high level and, importantly, the access prerequisite. Only the lack of differentiation from sibling candidate-listing tools keeps it from being complete.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, with page, job_id, and label_id all documented in the schema itself, so the baseline is 3. The description adds nothing beyond the schema — it does not clarify what an '申请列表' entry contains or how label_id's filter values interact with job_id.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description names a specific verb ('查看' / view) and a specific resource ('候选人投递申请列表' / list of candidate applications) and scopes it to recruiter mode. It does not explicitly differentiate itself from close siblings such as boss_hr_candidates or boss_hr_recommendations, which likely also list candidates, so it falls short of a 5.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability block ('roles=recruiter; recruiter_platforms=zhipin-recruiter') gives a real prerequisite that tells the agent when this tool is callable. However, there is no guidance on when to pick this over the many sibling list tools (boss_hr_candidates, boss_hr_recommendations, boss_hr_jobs), so the usage is only implied rather than directed.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description must carry the behavioral burden, and it does disclose the role/platform gate the agent needs before invoking. It stays silent on other behavior: nothing about what select=true does, pagination limits, or whether searching has any side effect on candidate records.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

It is a single short line, front-loaded with purpose before the bracketed availability metadata, so there is no wasted prose. The bracket notation is compact but slightly opaque compared to plain phrasing.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 12-parameter search tool with no output schema and no annotations, the definition covers eligibility but not the return shape, result volume, or pagination expectations. It is adequate to identify the tool but thin for a tool this configurable.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so every one of the 12 parameters is already documented in the schema; the baseline of 3 applies. The description adds no syntax, defaulting, or interaction guidance (e.g., how query interacts with the other filters) beyond what the schema states.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description names a specific verb (搜索/search) and resource (候选人/candidates) and scopes it to 招聘者模式 (recruiter mode), which separates it from the seeker-side boss_search without needing the schema. The bracketed availability metadata adds mode context but the phrasing is terse and partly encoded rather than explained.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The availability clause (roles=recruiter; recruiter_platforms=zhipin-recruiter) tells the agent when the tool is callable, which is a real precondition. However, it gives no guidance on when to prefer this over siblings such as boss_hr_recommendations or boss_hr_applications, and no exclusions beyond the role gate.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations supplied, the description carries the full burden; it does disclose that results include unread counts and recent message summaries, which signals a read-oriented listing. It does not state whether viewing marks messages as read, how many items are returned per page, or any rate/permission limits beyond the platform tag.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

The core purpose is front-loaded in a single compact clause, with the availability tag appended as machine-readable metadata. There is no redundant phrasing, though the bracketed availability string is somewhat dense.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 3-parameter, all-optional listing tool with full schema coverage and no output schema, the description partially compensates by naming the returned fields (unread count, last message summary). Pagination volume and read-state side effects remain unstated, keeping it just short of complete.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100% and all three parameters (page, job_id, label_id) already carry descriptions, including the label_id enum meaning (0=all, 1=new greetings, 2=in-progress). The description adds no syntax or default information beyond the schema, so the baseline 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb+resource: 查看与候选人的沟通列表 (view the chat/conversation list with candidates), and adds what the result contains (unread counts, latest message summary). This differentiates it from message-level siblings like boss_hr_chatmsg and boss_chatmsg, though it does not explicitly contrast with the similarly-named boss_hr_last_messages.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The '招聘者模式' prefix plus the availability tag (roles=recruiter, recruiter_platforms=zhipin-recruiter) clearly scope when the tool is applicable. However, there is no explicit guidance on when to prefer this conversation-list tool over siblings such as boss_hr_last_messages or boss_hr_chatmsg, so routing is implied rather than stated.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are supplied, so the description carries the full burden. It usefully discloses the role and platform gating, but says nothing about read-only semantics, pagination direction, or result ordering beyond what the max_msg_id parameter name implies.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

A single compact sentence with the role mode front-loaded and the availability constraint bracketed at the end. Nothing is wasted, though the bracketed metadata is dense and slightly cryptic.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a simple read-only history fetch with three fully documented parameters and no output schema, the description covers the essential access constraint. Only the paging behavior and return shape are left implicit, which is a minor gap.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so all three parameters are already documented in the schema. The description adds no extra meaning about friend_id sourcing, count limits, or how max_msg_id interacts with paging, so the baseline 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource ('查看与指定候选人的聊天消息历史') and prefixes it with '招聘者模式', which separates it from the candidate-side sibling boss_chatmsg. It does not explicitly name the sibling, but the role qualifier makes the distinction inferable.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The availability clause (roles=recruiter, recruiter_platforms=zhipin-recruiter) tells the agent when the tool is callable, which is real guidance. However, it never contrasts with boss_hr_chat, boss_hr_last_messages, or boss_chatmsg, so the agent must infer which chat-reading tool to pick.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the burden and usefully discloses that it will not automatically agree to anything and will not overwrite existing files. It also specifies role and platform availability, though it omits return behavior and permission details beyond that availability.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single front-loaded sentence with no wasted wording. The key operation and constraints are stated before the availability metadata.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a three-required-parameter download tool with full schema coverage, the description covers the essential operation, availability constraints, and important side-effect limits. It could say more about failure conditions or return behavior, but the schema already handles parameter meaning.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so the input schema already documents all three parameters thoroughly. The description reinforces that the resume must already be received and accessible, but adds little parameter-level meaning beyond the schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description states a specific verb and resource: downloading a candidate's attached resume that has been received and is accessible. It implies differentiation from related actions by clarifying that it does not auto-accept, but it does not explicitly name sibling alternatives.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

It establishes clear usage conditions: recruiter mode, resume already received, and access allowed. It also notes behavioral constraints (no auto-accept, no overwriting), though it does not explicitly compare against sibling tools like accept_resume or request_resume.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the full burden but only says this is a 'request' to exchange. It does not disclose whether the candidate must approve, whether the contact is revealed immediately or only after consent, whether the operation is reversible, or what rate/permission constraints apply.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

A single short, front-loaded sentence with no filler. The bracketed metadata is dense and machine-styled rather than prose, but it is compact and placed after the core action.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a two-parameter, no-output-schema, no-annotation outbound request tool, the definition covers the platform/mode gating but leaves the consent flow and result behavior unexplained. It is minimally adequate rather than complete.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100% and the type parameter already has an enum with a default, so structured data does the heavy lifting. The description's mention of phone/WeChat loosely maps to the type parameter but adds no format or behavior detail beyond the schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description names a specific action and resource: '请求交换候选人手机号或微信' (request exchange of candidate phone/WeChat), scoped to recruiter mode. It does not explicitly distinguish itself from the sibling boss_exchange, but the 'recruiter mode' framing gives an agent enough to know which side of the platform it applies to.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability clause (roles=recruiter, recruiter_platforms=zhipin-recruiter) tells the agent it is recruiter-only, which is real usage guidance. However, there is no statement of when to use this versus boss_exchange, boss_hr_request_resume, or boss_hr_chat, and no exclusion conditions are given.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description must carry the behavioral load, and it does disclose two meaningful traits: an explicit operator-approval gate and that no read receipt is sent. It omits failure modes, retry/idempotency behavior, and consequences of calling it on a candidate who is already in a session, leaving real gaps for a write-type tool.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

Front-loaded with the mode, action and the approval constraint; the availability metadata is compactly bracketed at the end. Every clause earns its place, though the bracketed platform syntax is dense and slightly hard to parse.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

Nine parameters, six required, a mutation with no annotations and no output schema: the definition covers purpose, approval, and one side effect, which is decent. It still leaves an agent unsure about the return value, what happens when the same geek_id/job_id is greeted twice, and whether `dry_run` output differs from a live run.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so all nine parameters, including the safety-relevant `yes` confirm flag and `dry_run`, are documented in the schema itself. The description adds no extra meaning about field formats or how the recommendation-card IDs relate to each other, so baseline 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

Names a specific action and scope: "招聘者模式：单次建立会话、发送首次招呼" tells the agent this is recruiter-mode, single-target session establishment plus first greeting. The "单次" wording implicitly distinguishes it from boss_batch_greet, but it never explicitly names that sibling or boss_greet as the alternative.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

States a precondition ("需操作者明确批准") and an availability gate (roles=recruiter, recruiter_platforms=zhipin-recruiter), which tells the agent when the tool is callable. It does not say when to prefer a different sibling, such as boss_batch_greet for multiple candidates or boss_hr_reply for continuing an existing conversation.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations at all, the description carries the full burden, yet it reveals almost no behavior beyond the availability line. It does not state that online/offline are state-changing mutations, whether taking a job offline is reversible, what confirmation or error behavior to expect, or whether page is ignored for those actions.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

A single, front-loaded sentence covers the operation and the availability constraint in a bracketed suffix. There is no redundant wording and the structure makes the two parts easy to scan.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

This is a no-annotation mutation tool with no output schema, so the description needed to explain the effects of online/offline and the shape of results. It only covers what the tool is and who can call it, leaving the behavioral consequences of the write actions entirely unspecified.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, so the schema already documents page, action and job_id, including that job_id is required for online/offline. The description adds nothing about parameters that the schema does not already say, which is the baseline 3 case.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description names a specific resource (job management list, 职位管理列表) and specific actions (list, online, offline, 上线/下线), which is enough for an agent to know what the tool does. It does not distinguish itself from the sibling boss_hr_jobs_detail, which covers adjacent territory, so it stops short of a 5.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability block (roles=recruiter, candidate_platforms=-, recruiter_platforms=zhipin-recruiter) tells the agent when the tool is applicable, which is genuinely useful gating context. However, it never says when to pick this tool over boss_hr_jobs_detail or when not to use it, so the guidance remains implicit.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

With no annotations, the description carries the full burden. It discloses the role/platform availability constraint and that the response includes the full postDescription, and 'view' implies a read-only operation. It does not state whether the call is side-effect free, how errors surface for an invalid enc_job_id, or whether permission/ownership checks apply.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

A single front-loaded clause states the action and the key payload, followed by a compact availability tag. No redundant sentences; the bracketed metadata is dense but earns its place as gating information.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a one-parameter read tool with no output schema, the description covers what is returned (full details including postDescription) and the access constraints. Only the error/empty-result behavior is unaddressed, which is a minor gap.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100% and there is only one parameter, so the schema already documents enc_job_id as the encrypted job ID. The description adds only that the job must be 'specified', providing no additional format, sourcing, or error semantics beyond the schema.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource: viewing the complete details of a specified job, and names the notable payload (postDescription). It does not explicitly contrast with the sibling boss_hr_jobs (presumably the list variant), so it earns a 4 rather than a 5.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability block tells the agent this only works in recruiter mode on the zhipin-recruiter platform, which is useful gating context. However, it gives no when-to-use guidance relative to alternatives (e.g. boss_hr_jobs for listing, or boss_ai_analyze_jd for analysis), so usage is only implied.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are supplied, so the description carries the whole burden of disclosure. It does establish the recruiter-only role and platform scope, which is real context, but it does not state whether this is a read-only operation, whether it consumes pagination from the visible chat list, or whether it marks conversations as read.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

It is a single front-loaded clause with the operating mode stated first, followed by a compact availability bracket. Nothing is redundant, though the bracketed metadata is terse to the point of being almost cryptic.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 4-parameter, all-optional listing tool with no output schema and no annotations, the definition covers availability but not what the message summary contains or how pagination terminates. It is adequate to attempt a call but leaves the agent guessing about the returned shape and about when to prefer a sibling.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%; every parameter (page, job_id, label_id, friend_ids) is documented in the schema, including the fallback behavior when friend_ids is omitted. The description adds no parameter meaning beyond that, so the baseline of 3 applies.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description names a specific verb and resource: '批量查看候选人最近消息摘要' (batch-view candidate recent message summaries) in recruiter mode. That is far more than the tool name alone conveys. However, it never distinguishes this from similarly named siblings such as boss_chat_summary, boss_hr_chatmsg, or boss_hr_chat, so the agent must infer which summary/chat tool to use.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability block ('roles=recruiter; recruiter_platforms=zhipin-recruiter') is a genuine precondition and useful for ruling the tool in or out. But there is no explicit 'use this when…' statement and no routing to alternatives for single-conversation message history or full chat content, so selection remains implied.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are provided, but the description identifies this as a read operation and discloses the required recruiter role and platform. It still does not clarify pagination behavior, whether any side effects occur, or the exact shape of the returned card data.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is a single front-loaded sentence with bracketed availability constraints and no wasted wording. The purpose and operating context are immediately clear.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a simple two-parameter read tool, the description covers purpose and availability, but it does not explain what a 'complete card' contains, how pagination works, or what the caller receives. With no output schema and no annotations, these gaps keep it at adequate rather than complete.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%, with both page and job_id documented directly in the schema. The description adds no parameter-specific meaning beyond the schema, so the baseline score of 3 is appropriate.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description states a specific verb and resource: recruiter mode reads recommended candidates' complete cards and parameters needed for first chat. It distinguishes the recruiter context from general recommendation/list tools, but does not explicitly contrast with siblings such as boss_hr_candidates or boss_hr_greet.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description gives role/platform availability constraints (roles=recruiter; recruiter_platforms=zhipin-recruiter), which implies recruiter-mode usage. However, it does not say when to use this tool instead of nearby alternatives like boss_hr_candidates, boss_hr_greet, or boss_recommend.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are supplied, so the description carries the full disclosure burden for what is an outbound, effectively irreversible action (a message delivered to a real candidate). It says nothing about delivery semantics, failure modes, whether prior threads must exist, or whether the message is sent immediately. Only the role/platform gate is disclosed.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

A single compact line with the action front-loaded and the availability constraint bracketed at the end. Nothing is wasted, though the bracketed metadata is dense and partially machine-syntax rather than prose.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

The schema documents both required parameters and the description supplies the role/platform gate the schema omits, which is the key missing context for a two-parameter reply tool with no output schema. It still lacks any note on outcome or how it differs from the other chat/reply siblings.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 100%: both message and friend_id are documented in the schema itself, so the description adds no parameter meaning. Baseline 3 is appropriate here since the schema does the work.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

"招聘者模式：回复候选人消息" gives a specific verb (reply) and resource (candidate messages) plus the operating mode (recruiter). It does not distinguish itself from adjacent siblings such as boss_hr_chat, boss_hr_chatmsg, or boss_ai_reply, so it stops short of a 5.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Does the description explain when to use this tool, when not to, or what alternatives exist?

The bracketed availability block (roles=recruiter; recruiter_platforms=zhipin-recruiter) implies the context in which this tool is valid, which is useful gating information. However, there is no explicit when-not guidance and no named alternative for reading a conversation versus replying into it, leaving the agent to infer routing.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations exist, so the description must carry the full behavioral load. It usefully discloses role and platform gating, which is the kind of auth/permission context the rubric credits. However, it says nothing about side effects (whether a message is sent into the candidate's chat), whether the request can be repeated or withdrawn, or what happens if the candidate never shares.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Is the description appropriately sized, front-loaded, and free of redundancy?

The definition is compact and places the action at the front, with constraints in one bracketed clause. The shorthand 'candidate_platforms=-' is opaque and takes a moment to decode, which slightly reduces clarity, but there is no filler text.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

This is a simple single-parameter tool, but it has no annotations and no output schema, so the description must explain what a call returns. It does not say whether the tool returns a request status or an error, or how the sharing result is later retrieved (via boss_hr_accept_resume or boss_hr_resume). That leaves an important gap for an agent that runs a multi-step workflow.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

There is only one parameter, and schema description coverage is 100%: friend_id is documented as '候选人会话 friend_id'. The description adds no additional semantics about which identifier to obtain or how to source it. A baseline of 3 is the correct result when the schema already carries the parameter documentation.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Does the description clearly state what the tool does and how it differs from similar tools?

The description uses a specific verb and resource: 'request candidate to share attachment resume' in recruiter mode, which is clear and unambiguous. It does not distinguish itself from closely related siblings such as boss_hr_accept_resume, boss_hr_download_resume, or boss_hr_resume, so it falls short of a 5.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.