propose_decision

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

Annotations only say readOnlyHint=false and destructiveHint=false; the description adds the crucial fact that the default path is non-mutating and that a commit flag flips it, which is real value beyond the annotations. It also discloses what the response contains (errors each carrying a `fix`, warnings, standing changes, impacted dependents, and the ADR markdown to save). It does not explain `acknowledge_impact` or what happens when commit=true fails validation.

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 dense sentences with the mode constraint front-loaded and the payload shapes given as inline examples, which is efficient for a structured parameter. The pseudo-code snippets are justified by the 0% schema coverage, though they make the prose slightly harder to scan.

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, yet the description still usefully summarizes returns, and it covers the two most complex parameters. The gap is `acknowledge_impact`, an 8th parameter that appears to gate the impact-warning path, plus `governs` — for a tool whose whole purpose is to surface consequences before a write, omitting how to acknowledge those consequences is a meaningful hole.

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 8 parameters, so the description carries real weight — and it does define the shape of `treats` (target, kind with four named values, clauses, reason) and `cites` (target, clauses), which the schema leaves as untyped objects. However, `governs`, `acknowledge_impact`, `body`, and `context` receive no explanation, leaving half the surface undocumented in both places.

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 opens with a specific verb and resource: review a draft decision against the graph, with the dry-run-versus-commit distinction stated in the same sentence. An agent can tell this apart from a plain read/query tool because the object is an uncommitted draft and the operation is validation rather than lookup.

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?

"before it lands" gives a clear temporal/contextual trigger, and "dry run unless commit=true" states the default mode explicitly. It stops short of naming alternatives (e.g., when to use this versus a preflight or status check) or stating when not to call it, so it falls just short of the top band.

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