redact_document

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

Beyond the annotations (readOnlyHint=false, non-destructive, non-idempotent), the description discloses a great deal: the copy is not uploaded and does not overwrite the source, the PDF is rasterized page-by-page removing text layer/form/attachments/metadata, small-text quality may degrade, interactivity is lost, repeated identical regions reuse the copy, and a changed copy file blocks reuse. It also notes a prior upload cannot be retracted. This is rich behavioral context beyond what annotations provide.

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

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

The core action and its non-destructive guarantee are front-loaded in the first clause, and every sentence carries substantive information (limits, side effects, reuse rules). It is dense and semicolon-heavy, but there is little filler to cut.

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 write operation with a 0%-coverage schema, this description is notably complete: it covers mutation semantics, side effects, irreversible prior-upload implications, reuse behavior, and downstream workflow. An output schema exists, so return-value detail is not required, and the mention of document_id is a bonus.

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 burden, and it does add meaning: regions are 1-100 normalized regions on the original pages. This clarifies the count constraint and coordinate normalization that the bare schema does not. document_id itself is self-evident, and box inner formatting is left to the schema, so not a full 5.

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+resource (locally redact regions of a document), specifies that it returns an independent copy document_id, and explicitly says it does not upload or overwrite the original. The privacy-redaction purpose ('遮盖指定姓名/学号等区域') is distinct enough that an agent can differentiate it from generic mutation 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?

It sets clear context: the agent must supply regions itself because PII is not auto-detected, and it warns to inspect all copy pages before uploading. It also directs downstream workflow (parse_document/read_evidence/save_study_record use the new ID) and reuse rules. However, it never explicitly contrasts with the closest siblings (crop_document, make_clean_copy) that might otherwise be chosen.

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