read_document

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

Annotations already cover readOnly/idempotent/non-destructive, yet the description adds substantial behavioral context: empty chunks do not mean blank or unanswered, Success/completed does not prove transcription completeness, review_hints are for locating images not error judgment, and format is textin_vlm_v4 without fabricated elements or confidence. This is exactly the value-add beyond 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 purpose is front-loaded and the content is dense but mostly earned. However, the same warning is repeated ('空分块不等于空白或未作答' and '无元素/分块不等于空白或漏答'), and the paragraph strings together many caveats without structure, reducing scannability.

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 the description correctly focuses on interpretation caveats (empty results, completeness of transcription, format differences) that an agent needs. It is nearly complete, missing only parameter-level guidance.

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% and the schema only exposes bare titles for job_id/limit/offset. The description implies pagination ('分页读取') but never explains the meaning, defaults, or units of limit/offset or what job_id must reference, so it fails to compensate 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 opening sentence '分页读取已缓存结果' (read cached results with pagination) states a specific verb+resource, and it names what each parse format returns (elements/page_summary for xParse, chunks/image_metadata for vlm_sync). It distinguishes itself from siblings by pointing to read_evidence and read_element for specific needs, though the heavy caveat load dilutes immediate clarity.

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 explicit routing: use read_evidence to verify images because VLM position isn't mapped to the original page, and use read_element for full chunks/image metadata. That is clear when-to-use-this-vs-alternative guidance. It lacks explicit exclusion conditions for the tool itself, but the sibling routing is genuinely helpful.

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