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

Annotations already declare readOnlyHint=false and destructiveHint=false, so the mutation-vs-read question is answered. The description adds context the annotations cannot: the report is delivered to a human team, it has no remediating side effect, and the note must avoid secrets — a real data-sensitivity constraint. It stops short of saying what the caller gets back (a confirmation, a ticket id?) or whether duplicate filings are deduplicated, which matters given idempotentHint=false.

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?

Three sentences, front-loaded with the core action before the constraint and the safety warning, with no filler. The failure-scenario list is slightly long but each item is a distinct valid trigger, so little would be gained by trimming it.

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 single-purpose reporting tool with no output schema, the definition covers purpose, selection rule, alternatives and a content restriction — enough to file a correct report. The remaining gap is the undescribed category enum and the unmentioned length limit on the note, which an agent would only discover from the raw schema.

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?

With 0% schema description coverage the description carries the full burden, and it does add crucial meaning the schema cannot express: exactly one of run_id, workflow_id, agent_id or build_id must be named, a mutual-exclusivity rule that is invisible in a flat object with no oneOf. It also frames what_went_wrong as prose in the reporter's own words. However, the category enum (wrong_result/failed/too_slow/too_expensive/confusing_tool/other) is never mentioned and the 2000-character cap on what_went_wrong goes unstated, so two of six parameters remain unexplained.

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 concrete verb and audience ('Tell the Fetch Hive team that something went wrong') and enumerates the triggering conditions: wrong result, failing workflow/agent, mismatched build, confusing tool. It explicitly rules out being a fix mechanism and names the siblings that do fix things (diagnose_run, edit_step, test run), so an agent can separate it from the remediation tools without opening a schema.

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 explicit when-to-use criteria (four concrete failure scenarios) and an explicit when-not-to-use with alternatives named ('it does not fix anything: to fix it now, use diagnose_run, edit_step and a test run'). It also adds a content restriction ('never put passwords, keys or personal data in the note') that governs how to fill the required field.

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