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

Annotations already declare readOnlyHint, idempotentHint and destructiveHint=false, and the description is consistent with a read-only deterministic compute node. Beyond the annotations it adds genuinely useful behaviour: the node is deterministic, execution happens server-side on Cloudflare Workers versus a browser delegation URL depending on compute mode and gpu flag, inputs are processed transiently and not stored or logged, an AP2 artifact with execution_hash is exported for chain provenance, and unknown policy keys are refused. It does not contradict 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?

The description is poorly front-loaded: it opens with name/title restatement and a long compute-mode paragraph that duplicates the compute parameter's schema description, then a full URL, then a 64-character FV-status hash URL, before finally stating the tool's actual job under 'Summary:'. The core purpose and output field list are crammed into the last two lines. Content is largely useful but the ordering wastes an agent's attention.

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?

With no output schema, the description compensates by enumerating the return fields (refused, reason_code, field, profile_id, profile_venue_label, as_of, contract_label, findings, evaluated_items, verdict, issues) and pointing to describe_tool for the full schema. It also covers provenance chaining, refusal semantics, and data-handling constraints. Remaining gaps are the shape of the nested terms/policy inputs and the meaning of verdict versus refused, which is minor for a node with rich structured fields.

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 all four parameters are already documented, including the compute enum and the parent_hashes/parent_tool_ids pairing. The description adds little semantic depth: it restates the required policy keys (profile, as_of, terms) that the schema already labels, and does not explain the expected shape of the nested 'terms' or 'policy' objects, which is the real ambiguity for a nested schema like this. Baseline 3 is appropriate.

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 eventually states a specific verb and object: lints a prediction-market event contract's declared settlement terms (source agency, fallbacks, revision rule, discretion, dispute window, outcome review, early close) against a caller-declared policy. That is far more specific than the name alone. It does not, however, distinguish itself from neighbouring tools such as analyze_prediction_market or validate_cat_bond_trigger_terms, and the actual purpose only appears in the final 'Summary:' sentence after several paragraphs of compute-mode boilerplate.

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?

There is no when-to-use or when-not-to-use guidance, and no alternative tool is named anywhere in the description. The only directive-like sentence is the privacy constraint 'Use synthetic or anonymised inputs only', which is an input constraint rather than a routing rule. An agent must infer from the tool name and the sibling list when this lint applies versus a general prediction-market analysis tool.

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