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

Annotations already declare readOnlyHint=false, destructiveHint=false, idempotentHint=false, so the description is not the sole safety source. It does add real context beyond that: recertify is a POST to /v1/capabilities/{id}/versions that publishes a replacement version with an engine declaration, and the body field set is enumerated. But it omits important behavior — what identifying/auth parameters (token, agent_key, delete_credential, payment_signature) mean here, whether recertify is reversible, and what status/pass_rates actually return.

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 description is a single unstructured run-on that interleaves business eligibility rules, three action-to-endpoint mappings, and a body field list in one block. Nothing is front-loaded cleanly (the action list starts mid-sentence with 'status:'), making it hard to scan and easy to misparse.

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 multi-action tool with 11 parameters, 9% schema coverage, no output schema, and a nested body object — a case that demands more description than average. The definition covers recertify's body reasonably but says nothing about what status or pass_rates return, how the 'house' field should be interpreted by a caller, or what token/agent_key are needed for, leaving the agent under-equipped for two of the three actions.

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 only 9% with 11 parameters, so the description must compensate. It does meaningfully document the nested `body` object for recertify (endpoint, protocol, input_schema, output_schema, machine_spec, price_credits, sla_seconds, engine, optional sla/effective_at) and ties `id`/`agent_key` to actions. However, seven parameters (token, handle, wanted_id, transfer_id, idempotency_key, delete_credential, payment_signature) are left entirely unexplained, which is a substantial gap in a low-coverage schema.

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 does name the domain (model compatibility/EOL: reading status and pass rates, or publishing a replacement version), but the opening clause is a dense run-on — 'Read model compatibility and pass-rate groups by lane, engine, and the house field ... after grace with at least five measured runs' — that buries the verb+resource under business rules. An agent can eventually infer it is a three-action EOL tool, but it isn't cleanly stated and the relation to siblings like capability_version_publish and capability_get is never drawn.

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?

The description lists what each action maps to as a REST endpoint but never says when to use status vs pass_rates vs recertify, nor when to prefer this over capability_version_publish or capability_get. There is also no mention of prerequisites (grace period, five measured runs) as a selection condition — it is only embedded as prose in the first sentence.

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