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

Annotations cover only readOnlyHint and openWorldHint; the description adds materially more: the rule-list basis (OWASP LLM01), the output verdict scale, the coverage limits (English-only, misses rewording), and the paid-call model ($0.002 USDC via x402, failed calls not charged) — a cost/auth constraint the agent must know before invoking.

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?

Three sentences, zero filler: what it does, what it returns plus the critical caveat, then price. The caveat and pricing are front-loaded rather than buried, and no sentence merely restates the name.

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?

Covers everything needed to call it correctly and interpret the result: input language assumption, verdict scale, match reporting, redacted-text return, known false-negative profile, and billing terms. With an output schema present, no further return-value documentation is required.

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 coverage is 100% with a single well-documented 'text' parameter (1–20000 chars), so the schema carries most of the load. The description still adds a meaningful constraint not in the schema — the input is expected to be English — which affects whether the result is trustworthy.

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?

States a specific verb and resource ('Screens one text for prompt-injection patterns') and enumerates the four pattern families it detects (instruction override, system-prompt extraction, jailbreak personas, hidden/encoded instructions). This is clearly distinguishable from nearby siblings like scam_phishing_message_filter or pii_credentials_redaction.

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?

Gives an explicit usage constraint and failure semantics: 'A rule list misses rewording and other languages: treat NONE as no known pattern, not as safe.' That tells the agent exactly how to interpret a negative result, which is the main misuse risk. It does not name alternative siblings to route to, so it falls short of a 5.

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