read_extraction

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

Annotations already declare readOnlyHint=true, destructiveHint=false, and openWorldHint=false, so the safety profile is covered. The description contributes meaningfully beyond that by disclosing the bounded/paginated nature of the read and the constraint on position fields, but it does not explain return format, pagination behavior, or cache freshness.

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

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

A single compact statement with the core operation front-loaded and the evidence caveat appended; there is no filler. Density is high, though the evidence clause bundles two ideas that could be structured more clearly.

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

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

With read-only annotations present and no output schema to explain, the description is adequate on operation and safety, but incomplete on selection context and on the three undocumented parameters. An agent knows roughly what it reads but not clearly when to prefer it over sibling readers.

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 largely does not. '有界读取' loosely implies the limit/offset window and '已缓存' implies job_id targets a cached job, but the defaults (limit=16000, offset=0), units, and job_id source are left entirely to the undocumented schema.

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

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

States a specific verb (读/read) and a specific resource (已缓存的TextIn原始抽取JSON文本), and the 'bounded' qualifier narrows the operation. It does not, however, distinguish itself explicitly from close siblings like read_parse_result, read_evidence, or read_document, so an agent must infer 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?

The description gives a downstream consumption rule ('fields without reliable positioning must not serve as verification-passing evidence'), but says nothing about when to choose this tool over read_parse_result/read_evidence/read_document. There is no when-to-use or when-not-to-use guidance for selection.

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