aiworkhub_manager_needfix_draft

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

Goes well beyond the readOnlyHint by disclosing that every returned field is mechanically derived (title from check id/summary, kind/severity from a fixed map, scope_files from changed paths, evidence from terminal_review.evidence, evidence_refs from the attempt-artifact manifest). It also states the description field is deliberately empty and that provenance is {origin: server_draft, verified: false}, so the agent knows exactly what is asserted vs. inferred. No contradiction with readOnlyHint=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?

The operational content is a strong opening and a strong closing sentence, but the middle is a long block of filing statistics (204 manual filings, 76 with accept results, 355 of 540 identifiers, p50 549 B, 1,796 archived cards). That evidence motivates the feature but does not help an agent select or invoke the tool, and it buries the 'file with needfix_add' instruction at the very end.

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 and readOnlyHint covers the safety profile, so the description's extra field-derivation detail is surplus rather than necessary. It is complete for a 3-parameter read-only draft tool. It does not cover the case where no evidence justifies any draft (empty result handling), which is the one behavioral question left open.

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 coverage is 0%, so the description has to carry the parameter burden — and it does: request_id delimits 'one request's own review evidence', while finding_id/check_id narrow to exactly one finding or one failing check. It does not state precedence or mutual exclusivity between finding_id and check_id, a minor remaining 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?

Opens with a specific verb+resource+scope: 'draft the NeedFixes one request's own review evidence justifies', and labels it READ-ONLY. It also separates itself from the write path by naming needfix_add. The differentiation is present but diluted by long statistical justification rather than a crisp 'this vs that' statement.

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?

Gives a concrete workflow: draft here, then 'file a candidate with needfix_add once you have written that description', and explains that finding_id/check_id narrow the draft to exactly one finding or check. It does not set exclusions against the other NeedFix creation siblings (needfix_convert, needfix_preview_convert), so it is clear context without explicit when-not rules.

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