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

Annotations already cover readOnly=false, destructive=false, idempotent=true, openWorld=true, so the safety profile is known. The description adds genuine behavioral detail beyond them: retry verifies without re-uploading, cancel does NOT delete CWA records, and resolve requires consistent language. These are useful operational constraints, though it doesn't explain repeated start behavior despite idempotentHint=true.

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 short and organized sentence-per-action, with no filler. But there is no front-loaded statement of what the tool is, and the dense mixed Chinese/English text with the undefined acronym 'CWA' hurts immediate comprehension.

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 annotations cover safety. The per-action behavior is largely covered, but the description omits the tool's overall purpose and any guidance for choosing acquire_book versus its four siblings, which matters for an openWorld multi-action 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 0% across 5 parameters, so the description must carry the burden and only partially does. It usefully maps parameters to actions (request_id/candidate_id/required_format for start; task_id/candidate_id for resolve), which the schema cannot express, but it never defines the meaning, expected values, or provenance of those ids.

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 precisely scopes each of the four actions (start creates a persistent task, retry only verifies submitted tasks, resolve selects a CWA match, cancel preserves records), so the agent knows what each invocation does. It also implicitly distinguishes itself from upload_book ('retry 不重新上传'). However, it never states the tool's overall purpose and uses an undefined term ('CWA'), so sibling differentiation against search_books/get_task is weak.

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 real routing context per action: retry is only for already-submitted tasks and does not re-upload, resolve is for explicitly choosing a match, and metadata differences after submission should be handled by re-searching rather than re-resolving. An exclusion ('不重新上传') and an alternative ('重新搜索再选择') are present, though it never addresses when to prefer acquire_book over the sibling tools.

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 (readOnly false, idempotent true, openWorld true, destructive false), so the bar is lower, yet the description still adds real value: in-MCP validation, automatic handoff of the same file to CWA, the fact that CWA-converted files are NOT delivered, and that unclear tickets or expired links do not re-consume quota. That quota and delivery-boundary context is not inferable from 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?

It is a single dense paragraph but front-loads the core purpose and proceeds through actions, parameters, and behaviors without filler. For a four-action tool it is appropriately sized, though the density of concerns in one block slightly reduces scanability.

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 still covers action semantics, parameter roles, CWA integration, quota behavior, and the delivery boundary. It is complete enough for correct invocation of a complex multi-action tool, with only sibling-level routing left implicit.

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 0%, so the description must carry the load, and it documents the roles of request_id, candidate_id, required_format, save_to_cwa (default true), and task_id per action. The action enum values (start/collect/retry/cancel) are each described. It could state format/type constraints more precisely, but coverage is otherwise strong.

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 opening sentence states a specific verb and resource (download EPUB/PDF original files from explicitly selected Z candidates) and adds in-MCP validation. It differentiates itself from siblings by describing the get_download_url handoff and the CWA integration. It loses a point only because the purpose is interleaved with per-action and parameter detail rather than crisply isolated.

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 explains the per-action conditions (start requires request_id/candidate_id/required_format; collect/retry/cancel use task_id) and when to use save_to_cwa:false for download-only. It routes the agent to get_download_url once get_task.download.status is available. It never explicitly contrasts with acquire_book or search_books, so the sibling routing is incomplete.

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 only readOnly and openWorld, and the description adds substantial behavior beyond that: auth requirements per mode (none for source, CWA Basic for opds, reachable file entry for gateway), expiry semantics (expires_at unknown=null, available_until bound by local reuse window and upstream expiry), the meaning of status:link_issued, that repeated calls do not re-fetch the ticket, and that verification is server-observed. It is consistent with the annotations and materially richer than the structured fields.

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 content is dense and mostly earns its place, but it is delivered as one semicolon-chained paragraph that opens with the delivery:source detail instead of a defining first sentence. It is hard to scan and not front-loaded, which hurts selection speed even though little is truly 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 fields need not be explained, and annotations already carry the safety profile. The description still covers auth, expiry, status meaning, and the client-side follow-up required. It is nearly complete for a 2-param tool, with only the task_id provenance left implicit.

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 0%, so the description must carry the meaning. It explains all three delivery enum values in operational terms (source=default original file link, opds=authenticated link, gateway=signed time-limited GET), which is the key disambiguation. task_id is only implied via 'download_book task' context rather than explicitly defined, leaving a small gap.

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 makes clear this tool resolves the downloadable URL for a task, with distinct delivery modes (source/opds/gateway). It never states its relationship to siblings like download_book or get_task as a routing rule, so an agent must infer when this is the right call versus those. Clear verb+resource, but no explicit 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?

It gives concrete when-to-use guidance per delivery mode: download_book tasks default to delivery:source, other tasks default to CWA-authenticated OPDS links, and delivery:gateway is an explicit request for a time-limited signed GET requiring a reachable file entry. It also notes the agent must still perform the HTTP save/read itself. No exclusion criteria relative to sibling tools, so not a full 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?

Annotations already declare readOnlyHint=true and openWorldHint=false, so the agent knows this is a safe, closed-world read. The description adds meaningful behavioral context by explaining what 'ready', 'available', 'needs_review', and 'blocked' states mean and that they don't imply the agent has saved the file. Still, it doesn't explain whether task state is persisted/immutable, what happens after retry, or any 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 dense run-on covering multiple state transitions and edge cases in a single paragraph with no headings or structure. It front-loads little and intersperses multiple concerns (ready, download.status, needs_review, blocked/waiting_for), making it hard to scan. It could be split into a purpose line and a status-reference list.

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?

Since an output schema exists, the description needn't explain return values, yet it spends most of its length interpreting status field values, which is useful but overlaps with what the output schema likely documents. It omits guidance on error cases (invalid task_id) and how this tool relates to its five siblings, leaving gaps for an agent navigating the 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?

With only one required parameter (task_id), schema coverage is 0% but the schema itself constrains the UUID format. The description adds no information about task_id beyond what the schema provides. Baseline of 3 is appropriate for a zero-description single-param tool, but no extra value is added.

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 clear purpose ('查询持久任务状态' = query persistent task status), correctly identifying a read of task state. However, the bulk of the text shifts to interpreting status semantics rather than clarifying what the tool returns or how it differs from siblings, leaving the core purpose somewhat buried.

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 explicit conditional guidance: when needs_review, choose choices; when blocked with waiting_for, handle the related task first, then retry to verify. This is actionable 'when to do what' context. It does not, however, name sibling tools or clarify when to use get_task versus get_download_url or download_book's status.

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, openWorldHint), and the description adds genuinely new context: the default source-priority order and the caveat that partial/warnings mean absence of results cannot be asserted. That last point is a meaningful, non-obvious behavioral trait.

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, front-loaded with the core purpose before the priority order and result-interpretation caveat. Dense and largely waste-free, though the source-priority ordering is somewhat 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?

An output schema exists, so return values need not be explained, and annotations carry the safety profile. However, with six undocumented parameters and no sibling routing guidance, the definition is only adequate for a tool of this complexity.

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 0% for all six parameters, so the description must carry the burden. It only loosely implies cursor behavior (next_cursor) and language preference (zh_preferred priority); author, formats, limit, and query semantics remain undocumented in both schema and 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 (搜索/search) and two concrete resources (书库 book library and 配置的书源 configured book sources). An agent can distinguish it from acquire_book/upload_book, though the description never names a sibling explicitly to sharpen the boundary.

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 (search before acquiring) but never stated. The description gives useful operating guidance about interpreting results (partial/warnings) and continuing with next_cursor, yet it provides no explicit when-to-use or when-not-to-use versus the sibling tools.

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 and idempotentHint=true, but the description adds genuinely new behavioral context: the two-step prepare-then-PUT protocol, that the returned link is where the full file goes, that refresh_ticket only updates authorization, and that files must never be passed through MCP. It stops short of covering failure modes, size limits, or ticket expiry.

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 dense and front-loads the 'prepare' action, with each semicolon-delimited clause carrying distinct information (params, metadata rule, PUT step, refresh action, prohibition). The mixed Chinese/English phrasing is slightly awkward for an English-only agent, but there is 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?

An output schema exists, so return values need not be explained, and annotations cover the safety profile. The description adequately describes the workflow and the file-transfer prohibition, but leaves sha256 and several parameter meanings undocumented for a 10-parameter tool, so it is 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?

Schema description coverage is 0%, so the description must carry parameter meaning, and it names most of them (request_id, title, filename, size_bytes, required_format, task_id, authors, language) plus the action values. It does not explain what sha256 is for, what request_id represents, or acceptable format values, so it only partially compensates for the coverage gap.

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 conveys a two-phase upload workflow ('prepare' to get a link, then PUT the file), so the resource and action are inferable. However, it never plainly states in a single clause that the tool uploads a book file to the library, and it makes no attempt to distinguish itself from siblings like acquire_book or get_download_url. Purpose is understandable but under-framed.

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 clearly routes between the two actions: 'prepare' lists the parameters to supply, while 'refresh_ticket' is used with task_id to refresh authorization. It also adds a hard usage constraint ('禁止通过 MCP 传文件或任意服务端路径'), which tells the agent what not to attempt. What's missing is any explicit comparison to sibling tools or a stated 'when not to use' condition.

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