build_evidence_pack

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

Beyond the annotations, the description discloses the most important behavioral trait: semantic=true sends the query to an external embedding service, which requires authorization and a pre-built index. This is real context an agent needs to avoid surprise failures. It does not explain why the tool is flagged readOnlyHint=false (what, if anything, is persisted), leaving one annotation unexplained.

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 tight sentences with the core capability front-loaded and the semantic-mode caveat second. Nothing is wasted, though the compactness comes partly at the cost of parameter coverage.

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, no-output-schema tool with 0% schema coverage, the description is thin: it omits the return shape of an evidence pack, how limit and char_budget bound the result, and the auth/index prerequisites are stated only as a consequence of the semantic flag rather than as prerequisites for calling the tool at all.

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 five parameters, so the description carries the full burden. It explains only semantic and hints at boundedness via char_budget; limit, project_id, query format, and the meaning of the char budget/limit interaction are left undocumented.

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+resource: building bounded evidence packs plus related reading clues. It is clear what the tool produces, but it never differentiates itself from the nearby read_evidence sibling or from search_library/index_semantic_search, so an agent has to infer its place in the family.

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 draws one useful decision boundary: default is local keyword search, while semantic=true routes to the project embedding service. That implies when to set the flag, but there is no explicit when-to-use-this-tool-versus-alternatives guidance and no warning about failure when authorization or an index is absent.

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