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

Annotations already cover the safety profile (readOnly, idempotent, non-destructive, closed-world), so the bar is lower, and the description still adds real context: it discloses the scoping rule that raw provider headers and bodies are never returned. That privacy/boundary statement is the kind of trait annotations can't express.

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?

One front-loaded sentence that leads with the action and then the return shape, with no padding. The 'looot:' prefix and the dense em-dash field list are minor noise but the whole thing is a single readable line.

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 no output schema, the description compensates by enumerating the returned fields and explicitly stating what is excluded, which is exactly what an agent needs to interpret results. The only remaining gap is the undocumented runId, whose semantics are left to the URL analogy.

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% and runId is never described in prose, but the reference to 'GET /v1/runs/{id}/attempts' implies the parameter identifies the run whose attempts are fetched, so the meaning is inferable rather than absent. A minimal baseline for a single-param tool with no schema documentation.

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 and resource ('get every attempt made for a run') and enumerates exactly what the response contains (status, provider response status, latency ms, receipt id, cost). It clearly reads as the per-attempt drill-down rather than the run-level tool, though it never names a sibling like runs_get to make the boundary explicit.

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?

There is no statement of when to reach for this tool versus runs_get or runs_list, and no prerequisites or exclusions. The only orienting clause is the API equivalence note ('same data as GET /v1/runs/{id}/attempts'), which is a mapping, not usage guidance.

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