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

Annotations only say readOnlyHint and openWorldHint; the description adds the substantive behavior the agent actually needs: it is a fixed rule list (deterministic, not model-based), it returns a four-level severity plus each match plus the text with matches removed, and it has a known blind spot for rewording and other languages. It also discloses the commercial terms — $0.002 USDC per successful call via x402 on Base/Solana/Arc, with failed calls not charged — which is critical pre-invocation information an agent cannot get from 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 definition is front-loaded with the verb and detection scope, then limitations, then pricing, in a logical order with no filler sentences. It is dense with parenthetical enumerations and runs slightly long, but every clause (categories, severity levels, limitations, cost) contributes information the agent needs.

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 single-parameter screening tool with an output schema, the description covers everything left open: detection method, category basis, return shape, explicit false-negative caveats, language restriction, and payment terms including what happens on failure. Nothing required to call it correctly or interpret its result is missing.

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% and the single 'text' parameter already documents the 1–20000 character bounds, so the schema carries the load. The description adds a modest constraint beyond the schema by specifying the text must be English and implying one text per call. That is a small increment over structured data, so the baseline of 3 applies.

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 gives a specific verb and resource ('Screens one text') plus the exact detection domain: instruction override, role/system-prompt extraction, jailbreak personas, hidden or encoded instructions, backed by OWASP LLM01 categories. This clearly separates it from adjacent siblings such as agent_intel_scam_phishing_message_filter or agent_intel_pii_credentials_redaction, which target different threat classes.

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?

'Screens one text' sets the scope (single input, not batch) and 'in English' sets the applicable language, which is real when-to-use guidance. The warning that 'A rule list misses rewording and other languages: treat NONE as no known pattern, not as safe' tells the agent how to interpret a negative result. It stops short of naming an alternative tool for non-English or semantic-level screening, so no explicit routing to a sibling is given.

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