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

Annotations already declare readOnlyHint=true and destructiveHint=false, and the description consistently reinforces this with the explicit 'never writes CRM or inventory, never replaces human authority' list, which adds meaningful context about side-effect boundaries. It also discloses behavioral traits beyond the annotations: the return shape ('eligibility or hold decision, a human action card, and an opaque receipt') and the restricted operator-only nature of two modes. No contradiction exists between description and annotations.

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 zero filler: the core function is front-loaded, followed by the return contract, then the safety boundary and mode entitlements. Every clause earns its place — the exclusion list is high-value disambiguation, not padding. The length is proportionate to the tool's complexity.

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 highly complex tool (6 params, deep nesting, output schema present), the description covers the essential ground: the evaluated rule set, the decision output, the audience restriction on modes, and a clear non-mutating boundary. The main gap is that it doesn't explicitly map each mode to its expected payload (evaluate→evaluation, record_stop→stop, register_governance→governance), though the schema's required fields and mode names largely carry that inference. Return values are covered partly by the description and partly by the existing output schema, so no deduction is needed there.

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 prose carries the burden for parameter meaning. The description partially compensates by decoding the critical required parameter, mode, explaining which modes customers are entitled to (evaluate, synthetic_acceptance) versus operator-only ones (register_governance, record_stop). However, the other five parameters (evaluation, stop, governance, tenant_id, fixture_id) receive no prose explanation, leaving the agent to infer mode-to-payload mappings from the schema alone.

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 names a specific verb ('checks'), a scoped resource ('a tenant-scoped lead'), and a concrete rule set (consent, STOP, quiet-hours, pacing, verified-sender, identity, assignment, manager-exception). The negative clause ('It never sends outreach...') sharply distinguishes it from send/action siblings, and the title reinforces the purpose. An agent can identify exactly what this tool does without opening the 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?

The description gives clear context for when to use the tool (evaluating a lead against compliance rules before follow-up) and explicit non-goals ('never sends outreach, clears STOP, writes CRM or inventory, reserves inventory, calculates a deal...'). It also partitions modes by audience ('entitled customer modes are evaluate and synthetic_acceptance; governance registration and STOP recording are operator-only'). It stops short of naming specific sibling substitutes for the excluded actions, so it earns a 4 rather than a 5.

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