cancel_job_run

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

Annotations already declare destructiveHint=true and idempotentHint=true, so the safety profile is partly covered. The description does add real behavioral value beyond that: the operation is POST rather than DELETE, a separate DELETE endpoint exists with different semantics, and the blanket convention rule means `confirm: true` is required in practice. However, it never says what actually happens to the run (does it abort in-flight work? is it reversible? what does the response contain?), which matters for a destructive mutation.

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 text is a research memo rather than a tool description: it leads with an 'EXPERIMENTAL' caveat, then a documentation-provenance justification for the `confirm` parameter. Sentences about 02-tools.md and 03-conventions.md earn nothing for the calling agent. Nothing is front-loaded — the core action is never stated at all.

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?

This is a destructive mutation with 0% schema coverage and no output schema, so the description should explain the effect of cancelling, the consequence of the run's state, and the role of `confirm`. Instead it documents its own authoring uncertainty, leaving the agent unable to predict outcomes. The 'EXPERIMENTAL/unverified' flag is useful honesty but does not substitute for operational detail.

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 four parameters, so the description must carry the load and largely does not. It shows project_id as a query parameter and the id/run_id placeholders in the URL, and it flags that `confirm` is required for consistency with the cancel_* convention — the one genuinely useful parameter fact. But job_id and run_id formats are never described, and the claim that confirm is a hard requirement conflicts with the schema, where it is optional with default false.

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 never states plainly that this tool stops a running job run; the agent must infer it from the endpoint path 'POST /v2/jobs/{id}/runs/{run_id}/cancel' and the aside 'presumably deletes the run record instead of stopping it'. It spends its opening words on provenance ('EXPERIMENTAL — from an unverified example request') rather than verb+resource. It does implicitly distinguish cancel from the separate DELETE endpoint, but not from the actual siblings (get_job_run, list_job_runs, create_job_run).

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 an agent should cancel a run versus polling get_job_run or listing runs, and no prerequisites or exclusions are given. The only routing-adjacent content is a note that a DELETE .../runs/{run_id} endpoint exists and is 'not exposed as its own tool in v1', which tells the agent nothing actionable about choosing between available tools.

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