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

Beyond the readOnlyHint/idempotentHint annotations, the description discloses crucial behavioral semantics: the distinction between could_not_verify (verification did not happen, with verification_error) and unsupported (no source covers it), the tolerance_pct override behavior, and the return of verbatim evidence with citations. These details tell the agent how to interpret results safely and are not available from annotations alone.

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 longer than typical but every sentence contributes: trigger phrases, use case, financial vs. other routing, return value summary, caller-important error semantics, and efficiency rationale. It is front-loaded with the most likely queries. Some redundancy exists (e.g., the phrase examples and 'replaces sequential calls' line could be trimmed), so it's not a perfect 5.

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?

The tool has no output schema, so the description must explain return values and edge cases, and it does: it enumerates the six verdict types, states that evidence and citation are included, and elaborates on the meaning of two failure states (could_not_verify vs unsupported). Given the tool's complexity, the description fully covers what the agent needs to invoke it and interpret results correctly.

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?

While the schema already documents both parameters (100% coverage), the description adds significant semantics: it clarifies what kind of natural-language claim fits the claim parameter, defines tolerance_pct's behavior (overrides claim wording, use 1–2 for hallucination detection), and explains the default cap. This goes well beyond the baseline 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 opens with clear trigger phrases and immediately identifies the tool's purpose: "natural-language claim verification against authoritative sources." It specifies the verb+resource (verifies claims against sources), and the examples of user phrasing make the intended use unmistakable and clearly distinguish it from sibling tools focused on search, research, and trading edge detection.

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 explicitly states when to use the tool: "Use whenever the agent needs to check whether something a user said is factually correct." It further provides routing policy for financial vs. non-financial claims and even notes that this replaces 4–6 sequential calls, giving a strong sense of when this is the right choice. However, it does not explicitly state when not to use it or name alternatives for exploratory or open-ended research, so it misses the top bar.

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