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

Beyond the readOnlyHint=true and destructiveHint=false annotations, the description discloses substantive behavioral traits: canonicality ('reads current server state and does not manufacture company facts'), explicit idempotency ('safe to repeat'), a confirmation boundary ('no additional confirmation is needed for this read'), non-execution semantics, and the fact that it never makes provider calls. It also warns that evidence completeness never satisfies human/external gates. No contradiction with 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?

The description is verbose (~180 words) but every sentence earns its place: purpose, negative space, evidence requirements, compliance disclaimers, prerequisite, canonicality, idempotency, and confirmation boundary. The compliance-gate enumeration ('NEVER completes KYB/KYC, terms, payout-bank...') is long but each item is a distinct gate an agent must not conflate with this tool's role. Core purpose is front-loaded.

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?

Given the high complexity — a deeply nested manifest schema, 3 parameters, and no output schema — the description covers the essential behavioral context: what it validates, what it cannot certify, prerequisites, and idempotency. The main gap is the absence of any description of the return value/report format, which the agent would need to interpret results since no output schema exists.

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 low (33%), so the description must compensate. It does add meaning for the checks parameter by enumerating the 12 control names that exactly match the enum, and it states that passed/failed results require the command and evidenceHash fields. However, the manifest parameter — the largest and most complex structure — receives almost no semantic explanation beyond the implicit 'caller-reported evidence' context, and _corply_context is entirely unaddressed.

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+resource: 'Inventory caller-reported local/sandbox evidence references against every required Corply Pay control,' followed by the exact list of controls. It then defines negative space ('does not execute or independently attest commands, establish sandbox or production readiness, or make provider calls'), which sharply distinguishes it from execution-oriented siblings like run_sandbox_payment_probe.

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?

Clear usage context is provided: the tool is for checking completeness of caller-supplied evidence, and explicit when-not boundaries are stated ('does not execute or independently attest commands...'). Prerequisites ('authenticated active organization access') and evidence-format requirements ('exact command and a SHA-256 evidence reference') are given. However, no alternative sibling tools are named by name, leaving the agent to infer which tool to use when execution or readiness attestation is actually needed.

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