audit_content

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

Annotations already mark it read-only, idempotent, and non-destructive. The description goes much further by disclosing the exact return shape (per-item pass/warn/fail verdicts, matched principle ids, concrete issues, before→after rewrite, aggregate summary), the deterministic heuristic rules per content type, and the offline execution model. There is no contradiction with the 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?

Every sentence carries information: purpose, return format, heuristics, offline behavior, and a sibling-pointer. The content is packed but not repetitive, and it is front-loaded with the core purpose before the details. Nothing is extraneous.

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 3 parameters, no output schema, and a large sibling set, but a call agent can invoke it correctly. The description covers the input domain, the heuristics, the return value, the offline behavior, and the alternative tool condition. Any missing details, such as array size limits, are nonessential for a read-only deterministic audit tool.

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 100%, so the baseline is 3. The description adds meaningful semantics on top by tying each `items.type` enum value to the heuristic that applies (e.g. metric must carry a number+unit, CTA/label action-led and ≤4 words, captions flag duplication of headings), which explains the purpose of the type field beyond its basic enum description. The optional `goals` and `system` parameters are already clearly said to be for traceability, so no further compensation is needed.

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 ('Evaluate an array of content items') and enumerates the accepted types. It differentiates itself from the closest sibling by stating it is used for per-item content verdicts rather than the principle library, so an agent can distinguish it clearly without inspecting schemas.

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 says 'Use this instead of evaluate_design when you need per-item content verdicts rather than the principle library,' providing a direct alternative and condition. It also signals 'Pure offline — no network or browser,' which helps an agent decide when the tool can be safely invoked.

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