Locus

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

Adds substantial behavior beyond the annotations: the call charges nothing, a 24-hour Stripe Checkout link is issued for agent-created accounts vs the billing page for a person's account, and the human must never be asked for card details. It also restates idempotency in operational terms ('the same idempotency_key returns the same link'), which concretizes the idempotentHint annotation rather than merely echoing it.

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?

Four short sentences, all front-loaded with the primary purpose and trigger first, and each earns its place except the 'human_message' instruction, which references a field the caller cannot see in the schema. Minor loss for that dangling reference.

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 an output schema present, return values need no explanation, and the description covers safety, cost, timing, and account-type branching well. The one gap is the unexplained 'human_message' hand-off: it is central to how the link reaches the human, yet it is absent from the schema and undefined in the description.

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 usd and idempotency_key are already documented, and the description adds only the idempotency-retry behavior already implied by the annotation. More concerning, it instructs 'Send human_message unchanged,' but no such parameter exists in the schema, which muddies rather than clarifies parameter meaning. Baseline 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?

States a specific verb and resource ('Get a payment link ... to add prepaid credits to this account') and immediately resolves the mismatch between the name add_credits and the actual behavior (returns a link, does not itself add credits). An agent can distinguish it from get_balance and the other siblings 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?

Explicit trigger conditions are given: 'Use when get_balance is low or a call failed for insufficient credits,' plus the follow-up step 'Check get_balance after they pay.' It names the sibling to consult both before and after, leaving nothing to 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?

Annotations already declare readOnly, idempotent, non-destructive, closed-world behavior, so the safety profile is covered. The description adds a cost trait not present in annotations ('Free') and frames the retrieve-before-execute workflow. It stops short of noting auth requirements or rate limits, but with annotations doing that work a 4 is warranted.

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, zero filler, with the primary action ('get one endpoint's full contract by slug') front-loaded and the workflow guidance following. Every clause carries 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?

An output schema exists, so return values need no explanation, and the description covers when to call it and how to get the slug. The only residual gap is that the five-parameter schema hints at search/pagination behavior that the description doesn't reconcile with the 'one endpoint' framing.

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 slug, model, model_cursor, operation_query, and operation_offset are all documented in the schema; the baseline is 3. The description reinforces that the lookup is keyed by slug but does not explain why model/operation_query appear on a single-endpoint describe call.

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 ('Get one endpoint's full contract by slug') and enumerates what the contract contains (input schema, example, output shape, price). This is clearly distinguishable from siblings like search_apis (discovery) and execute (invocation).

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?

Gives explicit timing conditions ('before the first execute or after an args rejection') and names where to obtain the required slug ('get the slug from search_apis'). The workflow relationship to execute and search_apis is unambiguous.

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?

Annotations already cover safety (readOnlyHint=false, destructiveHint=true, openWorldHint=true), and the description adds real value beyond them: it discloses that the call charges credits, that max_credits caps the charge, and that a stable idempotency_key yields exactly-once retries surfaced as idempotent_replay:true — a conditional idempotency the flat idempotentHint=false cannot express. It omits the approval_token/approval-error flow, which is the one notable behavior 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?

Four tight clauses, front-loaded with the core action and cost, followed by discovery, cost-cap, idempotency, and continuation. No filler and no repetition of 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?

With an output schema present, the description needn't explain return values, and it still flags the idempotent_replay field and the long-result handoff. Combined with 100% schema coverage it is nearly complete; the missing piece is any mention of the approval_token / approval-error path, which an agent will only discover from the schema.

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 all seven parameters; the description largely restates it (max_credits caps the charge, idempotency_key reused on retry). The only genuinely additive bit is the response field name idempotent_replay:true, which is output-side rather than parameter semantics. Baseline 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?

Specific verb+resource: 'Call one enabled endpoint by slug with args from its schema.' It immediately distinguishes itself from siblings by naming search_apis (for slug discovery) and get_call_result (for continuation). An agent can identify this as the single execution entry point 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?

Explicit routing: 'Get the slug from search_apis' tells the agent the prerequisite discovery step, and 'Long results continue via get_call_result' names the follow-up tool. It also states the cost-control condition ('max_credits caps the charge') and the retry condition for idempotency_key, so the when/how is fully specified.

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?

Annotations already cover the safety profile (readOnly, idempotent, non-destructive, closed-world), so the description only needs to add cost/auth context — and it does: 'Free' and 'secret key' auth for the end-user scope. It does not discuss rate limits or eventual consistency of the balance, which keeps it out of 5 territory.

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?

Three short clauses: what it reads, how scope is selected, when to call it, plus a one-word cost note. Zero filler, and the primary purpose is front-loaded ahead of the usage triggers.

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?

An output schema exists, so return-format details are unnecessary here. Between the description and schema an agent knows both scopes, the auth mode each requires, the trigger conditions, and that the call is free — nothing needed to invoke correctly is missing.

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 already documents external_user_id's type and auth requirement, so baseline is 3. The description goes beyond the schema by clarifying the default (omitting the parameter returns the pooled account balance rather than failing), which is behavioral meaning the schema does not encode.

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 ('Read the remaining credit balance') and immediately disambiguates the two scopes it can serve: the pooled account, or a single end user via external_user_id. An agent can distinguish it from the write-side sibling add_credits 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?

Gives two concrete triggers ('before a large spend', 'after an insufficient-credits error') that make the call condition actionable. It also states the auth precondition for the per-user mode. It stops short of naming an alternative tool or stating when not to use it, so it's strong but not exhaustive.

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?

The annotations already declare read-only, idempotent, non-destructive, and closed-world behavior. The description adds useful context beyond those hints by explaining chunked reading, offset/page handling, async waiting, and that the operation is free.

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 sentence is densely packed but front-loaded with the core action and receipt ID, then adds truncation, chunking, and wait behavior. Every clause carries operational value.

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 has an output schema, rich annotations, and fully documented parameters. The description supplies the missing usage context: when to call it, how chunking works, and what wait_seconds does.

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 still adds meaning by clarifying that results are read in chunks using offset or page and max_characters, and that wait_seconds controls waiting for a still-running call.

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: read a stored call result by receipt ID. It also distinguishes the tool from execute by saying to use it when execute truncated the result, so an agent can identify its role 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?

It gives a clear trigger condition: use when execute truncated the result, and explains that wait_seconds is for a still-running call. It does not state explicit exclusions or name alternative tools, but the operative context is unambiguous.

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?

Annotations already declare readOnly/idempotent/non-destructive/openWorld=false, so the safety profile is covered. The description adds non-obvious context the annotation can't: output is text, cost is 'Free', and there is a preferred skill-file alternative. It does not mention pagination behavior via cursor or size limits, hence not a 5.

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?

Three short sentences, no filler. The primary purpose and the alternative-skill condition are front-loaded before the parameter modes, so the most decision-relevant information comes first.

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?

An output schema exists, so return-value explanation is unnecessary, and all six parameters are documented in-schema. The description covers purpose, routing, and the skill fallback; only pagination/chunking behavior (cursor + max_characters interaction) is left to the schema, 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 every parameter has a terse description ('Omit to list or pick by task', 'Your goal, to pick a guide', 'Prior next_cursor', etc.), so the schema carries the load. The description's id/task routing rule is genuinely useful but overlaps with the usage-guidance dimension, and it says nothing about cursor, section, version, or max_characters. Baseline 3.

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?

Specific verb + resource + format: 'Fetch a Locus how-to guide as text.' It also states scope (how-to guides, not API docs) and is easily distinguished from siblings like describe_api or search_apis.

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?

Explicitly states the when/when-not condition: use this 'when no Locus skill files are installed; otherwise prefer the skill.' It also gives mode selection (omit id to list, pass task for a goal-targeted guide), so an agent knows exactly which call shape to use.

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?

Annotations already declare readOnlyHint, idempotentHint and destructiveHint=false, so the safety profile is covered. The description adds genuinely new behavioral context: results are newest-first, the call is free (no credit cost), and it is paged. It does not quantify page size or cursor expiry, so not a 5.

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 dense sentences with zero filler; the what/ordering/fields come first, then usage and filters. Every clause carries information the agent needs.

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?

An output schema exists, so return-value documentation is not the description's burden, yet it still names the key fields. Combined with the paging and free-operation notes and the sibling routing, an agent has everything needed to call this correctly.

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 are documented in the schema and the baseline is 3. The description only restates that status and provider act as filters, adding intent but no syntax or format detail beyond the schema's enum and default notes.

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 ('List recent billed calls'), gives the ordering ('newest first'), and enumerates the returned fields (call ID, endpoint, status, credits, error summary). It also names the sibling tools that consume the resulting ID, so an agent can distinguish it from get_call_result and report_bug without opening a 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?

Explicitly names two use cases ('audit spend' and 'find an ID for get_call_result or report_bug') and the filter axes (status, provider), which is clear routing guidance. It stops short of stating when NOT to use it or what the alternative is for a full-text or user-scoped search, so it falls just short of the top band.

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?

Annotations already declare readOnlyHint, idempotentHint and destructiveHint=false, so the safety profile is covered. The description still adds real context beyond them: 'listing changes nothing', that the operation is 'Free', and that results are 'paged'. It does not describe pagination mechanics or rate limits, keeping it short of a 5.

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 what the resource list is, then the action guidance, then cost/paging in three clipped fragments. Dense but every clause carries information; the trailing 'Free, paged.' is terse but purposeful.

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 an output schema present, return values need no explanation, and the description still adds useful field names, cost, and pagination behavior. It also points to the mutation endpoints for release/renew, so an agent has everything needed to call and act on this 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?

Schema description coverage is 100%, so cursor, status, service and end_user are already documented in the schema, and the description adds no filtering syntax or defaults beyond that. The mention of 'service, status, and expiry' describes returned columns rather than input parameters, 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?

Names a specific verb (List) and a precise resource class (durable resources created by your calls that persist, expire, or cost money), reinforced with concrete examples (AgentMail inboxes, phone numbers, E2B sandboxes) and the fields returned. An agent can distinguish this from list_calls or get_balance 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?

States the decision context clearly: use it to find what to release or renew, and even names the follow-up endpoints (POST /credits/resources/{id}/release or /renew). It also rules out side effects with 'listing changes nothing'. No explicit when-not-to-use or sibling alternative is given, but the routing intent is unambiguous.

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?

Annotations already declare readOnly, idempotent, and non-destructive. The description adds genuinely new context beyond those: the operation is free, and it is purely observational with no access-granting or execution side effects. It does not cover pagination behavior, but the output schema handles the return shape.

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?

Three short clauses, front-loaded with the purpose before the parenthetical disambiguation and the negative boundaries. Every sentence earns its place 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 read-only list tool with full schema coverage, rich annotations, and an output schema, the description supplies everything an agent needs: purpose, orientation role in the workflow, and explicit side-effect boundaries.

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 limit, query, cursor, and group_type are already fully documented in the schema. The description adds no parameter-level meaning beyond that, which is the expected baseline when the schema does the heavy lifting.

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 (browse) with a specific resource (catalog categories and packs), and defines what a pack is ('curated endpoint bundles'). It explicitly names the sibling tools it is not: search_apis, request_tool_access, execute.

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 when to use it ('to orient before search_apis') and draws explicit negative boundaries: it enables nothing (use request_tool_access) and calls nothing (use execute). Both the condition and the alternatives 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 annotations only covering the safety profile, the description adds substantial context: persistence, workspace-wide scope, the context-cost consequence on every session, and the first-pin acknowledge_context_cost:true prerequisite. These are behavioral traits not derivable from the annotations.

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?

Four dense sentences, front-loaded with the core action and default, then scope, then the prerequisite. No filler and nothing repeated.

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?

An output schema exists so return values need no explanation, and the description covers mutation semantics, scope, and the one-time acknowledgement prerequisite. Nothing needed to invoke it correctly is missing.

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 all three parameters. The description reinforces the acknowledge_context_cost requirement and the pinned default, but adds little syntax or meaning beyond the schema's own descriptions.

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 (pin/unpin), the resource (one enabled endpoint), and the precise effect (appears as its own typed tool on every MCP connection instead of going through execute). This clearly differentiates it from the execute sibling.

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?

Gives a clear when-to-use rule ('Use for endpoints you call repeatedly') and implicitly routes one-off calls to execute. It also states what the tool does not do ('does not enable or call the endpoint'), though it never names a negative alternative explicitly.

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?

Goes well beyond the annotations (which only say non-readOnly, non-idempotent, non-destructive) by disclosing cost ('charges nothing'), the absence of retry/refund side effects, and that no in-session reply is returned. These are exactly the behavioral traits an agent needs before invoking a fire-and-forget reporting 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?

A single dense sentence: trigger, action, required traceability fields, and behavioral caveats, all front-loaded with zero 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?

An output schema exists so return values need no explanation, and the description pre-empts the main agent confusion ('no in-session reply') for a non-readOnly, non-idempotent reporting tool. Nothing material is missing for correct invocation.

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 all four parameters. The description only adds the motivation for call_id and tool ('so it can be traced') and says nothing about the message or context payload. Baseline 3 is appropriate when the schema carries the load.

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 ('File a bug report with the Locus team') plus the triggering condition ('after a wrong result or error'). It is clearly distinguishable from siblings like get_call_result, list_calls, or execute, none of which file reports.

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?

Gives a clear when-to-use trigger (after a wrong result or error) and an implied not-to-use for normal successful flows. It does not explicitly name a sibling alternative or exclusion, so it stops short of a 5.

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?

Beyond the annotations (which only say non-read-only, non-destructive, non-idempotent, closed-world), the description discloses that the grant is short-lived, that approval is human-mediated in the normal case, and that an unclaimed agent-created account enables immediately. 'Calls nothing; free' rules out upstream invocation and billing side effects, which is genuinely useful context.

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 dense sentences with the primary case front-loaded and the exception second; no filler. The telegraphic clauses ('status enabled', 'Calls nothing; free') are terse but readable, and nothing appears wasted.

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?

An output schema exists, so return values need not be explained, and the description covers both outcome branches of calling it. Minor gaps remain: behavior when the endpoint is already enabled, or what happens when the link expires or is denied.

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 the exact provider/endpoint slug pattern and the 'reason' field (shown to the human) are already documented in the schema. The description reinforces that exactly one endpoint is targeted but adds no syntax or format 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 states a concrete verb and resource: creating a short-lived approval link for a single disabled endpoint, plus the alternate auto-enable path. It is clearly distinct from execution-oriented siblings like execute or run_task, though it never names any sibling 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 by the scoping constraint ('one disabled (enabled:false) endpoint'), so an agent can infer this is the path when an endpoint is unavailable. It gives no explicit when-not guidance and does not point at alternatives such as execute, describe_api, or search_apis for other cases.

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?

Annotations only declare readOnlyHint=false, openWorldHint=true and idempotentHint=false; the description adds the reason for the non-read-only flag (billed per fulfilled provider call), a concrete cost estimate (~10.1 credits), how mode drives provider breadth and cost, how maxCredits caps spend, and that the result reports credits charged. It also flags rails/strategy as legacy, which structured fields cannot express. Minor gap: the schema exposes an idempotency_key while idempotentHint=false, and the description does not reconcile that for retry behavior.

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 front-loaded sentences; purpose comes first, then alternatives, then billing. The second sentence is dense with semicolon-chained clauses (billing, mode, maxCredits, legacy fields, cost estimate), which is efficient but slightly harder to scan than a short bulleted flow.

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 17-parameter, cost-bearing tool with an output schema, the description covers what the output schema cannot: routing versus siblings, spend semantics, and deprecation of legacy fields. It leaves the geo/recency/domain-filter parameters and the idempotency question unaddressed, so it is strong but not fully 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?

With 17 parameters and only 41% schema description coverage, the description compensates for the highest-risk fields: it explains mode's fast/balanced/comprehensive trade-off between provider breadth and cost, that maxCredits is a total ceiling, and that rails/strategy are legacy with 'prefer mode'. It says nothing about topic, country, recencyDays, maxResults, maxProviders, includeImages, include/excludeDomains, providers, attribution, or externalUserId, which remain undocumented in both places.

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 (search the live web) plus the exact return shape (ranked, deduplicated sources with title, URL, snippet merged from several providers) and explicitly notes no LLM summarization on this path. That negative clause alone separates it from web_research and web_extract.

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?

Gives an explicit when: 'use when you need links to cite or read next', and names two alternatives with the condition that selects each (web-research for a synthesized answer, web-extract to read URLs). Nothing is left to 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?

Annotations mark this as a non-read-only, open-world operation; the description adds substantive context beyond that: it spends this account's credits under spending controls, verifies results, and returns sources and cost. The async contract (status running → call again with action 'wait' and task_id) is disclosed, which annotations do not cover. It stops short of describing failure/partial-failure behavior or whether a cancelled task still incurs charges.

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 routing rule and cost/verification behavior are front-loaded, and the trailing sentence about 'wait' is actionable rather than filler. The two inline examples are slightly verbose but earn their place by demonstrating the expected goal 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 10-parameter, nested-schema, open-world orchestration tool with an output schema, this covers the essentials an agent needs: when to choose it, how to phrase the goal, what it costs, and how to handle the asynchronous lifecycle. Return value details are correctly left to the output schema.

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 defines each parameter, but the description adds real meaning: 'goal' is described as plain language with two concrete example phrasings, and the submit/wait/task_id interplay is explained procedurally rather than just field-by-field. It does not elaborate on how max_credits interacts with the default or how deadline_seconds affects partial delivery.

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 scope: it finds and calls the right APIs to accomplish a multi-step goal, returning structured data, files, sources, and cost. It explicitly contrasts with the sibling 'execute' for single calls, so an agent can route correctly without opening both schemas.

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?

Gives an explicit precedence rule ('Use this first whenever a request needs more than one API call'), enumerates the qualifying scenarios (lists, enrichment, multi-source research, filings, CSV deliverables), and names the alternative ('Use execute only for a single call'). It also tells the agent what to do on a running status instead of resubmitting.

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?

The annotations already declare read-only, idempotent, non-destructive and closed-world behavior, so the bar is lower, yet the description still adds cost ("Free"), result shape (slug, price, enabled state), and ranking behavior (enabled first). Nothing here contradicts the annotations.

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 routing advice and chaining instructions are front-loaded and every clause is actionable. The long domain enumeration is dense and slightly list-like, but each domain keyword supports semantic matching, so it largely earns its space.

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 full schema coverage, an output schema, and complete annotations, the description only needs to supply selection and sequencing context, which it does. Chaining to describe_api/execute/run_task and the free/ranked return shape make the definition self-sufficient.

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 documented parameters set the baseline at 3, but the description adds meaning beyond it by explaining that queries should be phrased as outcomes ("twitter user followers") rather than provider names. It says nothing extra about pack, category, cursor, or include_facets, leaving those 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?

The description opens with a concrete verb-plus-resource ("Find a live data API") and enumerates the domains it covers, so an agent can tell instantly whether this tool serves its need. It also differentiates itself from "scrape with curl", "answering from memory", and from siblings describe_api, execute, and run_task.

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 when to use it (before scraping or answering from memory), what to do next (describe_api for schema, then execute), and which alternative to pick when an outcome spans multiple calls (run_task). That is explicit when-to-use, when-not, and alternative 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?

Annotations already declare the safety profile (readOnly, idempotent, non-destructive). The description adds meaningful extra context: it renders a panel to the human rather than returning machine data, and it is 'free,' which matters because several siblings (execute, web_search) incur cost.

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 tightly packed sentences, front-loaded with the core action and contents, followed immediately by the when-to-use clause. The section enumeration is long but each item earns its place by telling the agent what the panel surfaces.

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 an output schema present and annotations covering safety, the description needn't explain returns. It covers contents, trigger conditions, and cost, leaving only sibling-alternative routing unaddressed.

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?

One optional enum parameter at 100% schema coverage ('Default activity.'), so the schema already carries the semantics. The description's section list loosely maps to the tabs but adds no format or default details beyond the schema, making baseline 3 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?

States a specific verb and resource ('Show the human an account panel') and enumerates exactly what the panel contains, making it clearly distinct from data-returning siblings like get_balance and list_calls — this renders a UI surface rather than returning raw data.

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?

Explicit when-to-use is given: 'Use when they ask to see their account, usage, or approvals.' It does not name alternatives (e.g., get_balance for programmatic reads), so the boundary against overlapping siblings is left to 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?

Goes well beyond the annotations: discloses per-URL billing, that maxCredits is a hard aggregate ceiling, the retry asymmetry (text mode retries only failed URLs with a second extractor, json mode uses a single extractor with no retry), the typical ~1.01 credit cost, and that the result reports credits charged. This is exactly the cost/retry information an agent needs before spending credits, and nothing contradicts readOnlyHint=false / idempotentHint=false.

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 core action, then usage routing, then billing and retry semantics in descending importance. Dense but every sentence carries operational information; the cost figure and retry clause could be tightened 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?

An output schema exists, so return shape needs no explanation, and the description still covers the non-obvious pieces: spend limits, retry behavior by mode, and cost per call. Only edge details such as failure/error outcomes for individual URLs are left unstated.

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 89%, so the schema already explains json, urls, maxCredits, maxAgeSeconds and the Locus controls. The description adds only the 1–20 URL bound and that maxCredits caps the total, which is largely redundant with the schema's minItems/maxItems and its own description.

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 the full content of one to twenty exact public URLs') and names the two output shapes (page text or structured JSON via json). The 'never searches or substitutes other pages' clause separates it cleanly from the router search/research siblings.

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?

Gives an explicit trigger ('Use after a search produced URLs or when the user supplies links') and an explicit alternative for the other case ('use router/web-search or router/web-research to find sources first'). The boundary between extracting and discovering is stated rather than 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?

Annotations already declare readOnlyHint=false, destructiveHint=false, openWorldHint=true and idempotentHint=false, so safety is covered. The description adds genuinely valuable non-schema behavior: the ~10.1 credit cost per call, the fixed-price nature of the light tier, that the result reports credits charged, and the legacy-path trap. However, it discloses nothing about how charges are incurred on failure or partial results, which matters for a metered, non-idempotent call.

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 outcome, then routing, then cost and the legacy-path constraint – all in roughly 90 words with no filler. The cost/dependency rule is compressed into one dense sentence that an agent may need to reread, but nothing is wasted.

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?

An output schema exists, so return-value explanation is rightly omitted, and the description still flags the credits-charged field. For a 19-parameter tool it covers the highest-stakes parameters (effort, maxCredits, filter fields) but says nothing about topic, recencyDays, maxResults, providers, outputSchema's effort dependency, or the mode/strategy fields.

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?

With only 47% schema coverage across 19 parameters, the description earns credit for explaining that effort controls depth and cost (light default, fixed price), that maxCredits is a hard ceiling, and that 'search filters' such as includeDomains also bypass the legacy path. It leaves the mode enum (fast/balanced/comprehensive) and the legacy strategy field unexplained, which is a real gap given they overlap conceptually with effort.

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 first sentence names a specific verb and deliverable ('Get a grounded, current answer with inline source references') and explains the mechanism (searches, reads sources, synthesizes). It explicitly separates itself from router/web-search ('sources only') and router/web-extract ('read known URLs'), so an agent can route without opening a 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?

Gives an explicit when-to-use condition ('when you need an answer, not a list of links') and names two alternative tools with their selection criteria. It also adds a critical operational warning: without effort plus maxCredits or search filters, the call silently falls back to a sources-only result with no answer.

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?

Annotations cover the safety profile (readOnlyHint=false, destructiveHint=false, idempotentHint=false), and the description adds genuinely useful context beyond them: no action runs code or charges credits, and save publishes exactly one draft revision that must already be validated. It stops short of clarifying archive reversibility or required permissions, hence a 4 rather than 5.

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?

Four sentences, front-loaded with purpose then routing, then the save precondition, then the side-effect disclaimer. Every sentence earns its place with no 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 14-parameter, 7-action tool with nested objects, the description adequately covers purpose, sibling routing, and the key save precondition, and an output schema exists so return values need not be described. It does not map which parameters apply to which of the seven actions, a modest remaining 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%, so every parameter is already documented in the schema. The description only lightly reinforces semantics ('save publishes one exact draft revision' maps to revision), adding little beyond structured data. Baseline 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?

States a specific resource (customer-owned Workflow code) and enumerates the exact lifecycle verbs (list, create, update, get, clone, archive, save). It explicitly distinguishes itself from workflow_validate and workflow_run, so an agent can route 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?

Provides explicit when-to-use and when-not-to-use: 'Use for authoring and versioning only; use workflow_validate to check code and workflow_run to execute it.' It also states the precondition that save requires the revision to have passed validation, leaving nothing to 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?

Annotations already declare idempotentHint and destructiveHint, so the bar is lower, and the description usefully goes beyond them: it explains that a repeated idempotency_key replays the existing run (replay:true) rather than starting another, and that max_charge_credits is a hard ceiling for the whole run. It does not cover auth/permission requirements or what state changes occur, keeping it short of a 5.

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?

Four dense sentences, each earning its place: purpose first, then mode semantics, then the credit cap, then idempotency and follow-up routing. 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 9-parameter mutation tool with an output schema, the description supplies the operational context an agent needs: mode choice, cost ceiling, exactly-once retry behavior, and where to go next for run lifecycle operations. Return-value explanation is unnecessary given the output schema, and nothing material is missing.

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, but the description adds real meaning beyond the schema: it defines pilot vs run semantics, frames max_charge_credits as a hard run-wide ceiling, and clarifies the replay behavior of idempotency_key. It adds value on the semantically tricky parameters rather than merely restating them.

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 ('Queue a run of one exact saved Workflow version'), and the emphasis on 'exact saved version' distinguishes it from siblings like workflow_validate or workflow_definition. An agent can identify the operation 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?

Explicitly explains the pilot-vs-run selection condition and routes follow-up work (progress, results, cancel, resume) to workflow_runs, which is a clear alternative. It stops short of an explicit when-not to use it (e.g., validating a definition first), so it is strong but not fully exhaustive.

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?

Adds context well beyond the annotations: call/artifact paging happens 'without a new charge' (cost behavior), get returns a 'bounded status', cancel only affects 'pending dispatch' (scope of the destructive operation), and resume needs 'current authorization'. These disclosures are non-obvious and directly useful for safe invocation.

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 sentences, zero filler, with the action inventory front-loaded and the routing rule in the closing clause. Every phrase carries distinct 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?

With an output schema present and annotations covering safety, the description supplies the remaining essentials: all six actions, their scope, cost behavior, and the sibling routing rule. Nothing an agent needs to select or invoke this correctly is missing.

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 every parameter, including run_id/call_id/artifact_id requirements. The description adds only marginal tying of call_id and artifact_id to the call/artifact actions, 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 verb pair (inspect/control) and the exact resource class (runs started by workflow_run), then enumerates all six actions in verb+object form that map to the action enum. It explicitly separates itself from the sibling workflow_run ('use workflow_run to start one').

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?

Gives an explicit lifecycle boundary: 'Use for everything after a run is queued; use workflow_run to start one.' The alternative is named and the selecting condition is stated, and eligibility constraints are noted for resume ('eligible paused run with current authorization').

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?

Adds real context beyond the annotations: no provider calls means nothing is charged, and fixtures run in an isolated no-network environment. This is consistent with readOnlyHint=false, since a successful inline check evidently produces a source_artifact_id. It does not cover idempotency or repeat-run behavior, so it stops short of a 5.

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?

Three tightly packed sentences, front-loaded with the core action and followed immediately by mode semantics, prerequisites, and cost. No filler or restated title 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 five-parameter tool with an output schema, the description supplies everything an agent needs that structured fields do not: the prerequisite ordering relative to save and run, the mode-dependent identity requirements, network isolation, and absence of charges.

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, but the description adds genuine meaning: 'mode check validates inline source (before it is persisted) or one exact draft revision; mode fixtures runs that revision against isolated, no-network fixtures' explains what the enum values actually do and the exclusivity of source vs. workflow_id+revision.

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 first sentence gives a specific verb and resource (check Workflow code) plus the key scoping distinction: without a production run. It then differentiates the two modes concretely, so an agent can distinguish this from workflow_run and workflow_definition 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?

Explicit routing: 'Use before workflow_definition save, which requires a passing check, and before workflow_run.' The two modes are each given a when-to-use condition (inline/pre-persistence vs. exact draft revision with fixtures), leaving nothing to inference.

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