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

Annotations already cover the safety profile (readOnly, idempotent, non-destructive, closed-world), but the description adds meaningful traits: deterministic computation, server-vs-browser delegation semantics, transient processing with no storage or logging, the 'synthetic or anonymised inputs only' restriction, and AP2 artifact export with execution_hash for provenance. These are genuine operational constraints an agent would not learn from the annotations. Minor gap: no indication of latency, quotas, or kernel-availability failures.

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 most important content — what the tool actually scores and what it returns — is the last sentence, behind several sentences of graph-plumbing boilerplate, a URL, and a hash receipt. Multiple sentences (FV-status snapshot, 'open at' URL) do not help an agent decide or invoke. It is not concise and it is back-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?

For a nested-parameter scoring node with no output schema, the description supplies the output field list (ccy, composite_grade, eligible, gaps, grade, home_regime, reserve_score, risk_score, tech_score, verdict), the input homogeneity rule, the provenance chaining in and out, and the compute-mode contract. What's missing is what shape issuer_profile expects and what the gap semantics or verdict thresholds are, which for a decision function is a notable omission.

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 schema already explains compute modes, parent_hashes/parent_tool_ids pairing and the policy_parameters container with its required issuer_profile. The description restates the compute-mode behavior and chaining, adding no syntax or format detail beyond the schema. Baseline 3 is appropriate; the description does note that unknown keys are refused only indirectly.

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 Summary sentence names a specific resource and action: scoring a non-USD stablecoin issuer's readiness to join Circle Partner Stablecoins on Arc against technical/operational, reserve-management and risk-management standards, and it enumerates the returned fields. That is well beyond a restated name. It loses a point because the purpose is buried at the end of a long boilerplate preamble, and it never distinguishes itself from arc-related siblings such as run_arc_fit_diagnostic or route_partner_stablecoin_jurisdiction.

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 chaining context (consumes art-42-arc-fit-diagnostic, feeds art-45-arc-xreserve-linter) and the synthetic-inputs-only constraint, which implies workflow placement. However, it states no explicit when-to-use or when-not-to-use condition and never names an alternative scoring tool, so the agent must infer selection from sibling names alone.

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