Safety Scrub

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

Annotations declare readOnlyHint=false, openWorldHint=true, destructiveHint=false. The description adds useful context: 'Best-effort, not a guarantee' (honest limitation), 'Pure pattern matching, no AI call' (no side effects/API cost), and '(1 MESH/call)' cost estimate. It does not fully detail edge behaviors like what happens with borderline patterns or performance on large text, but with annotations carrying the safety profile, this is adequate.

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 compact and well-front-loaded, with the core purpose in the first clause. It packs the data-type list, output contract, limitations, and cost into a dense but readable block. Slightly crowded with the trailing meta fields ('a tool · safety', '1 MESH/call'), but no wasted words overall.

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-param, no-output-schema tool with 100% param coverage, the description covers the essentials: what data types get scrubbed, the output shape, and the best-effort caveat. The lack of output schema means the return-format explanation (scrubbed/redacted/kinds with example) is genuinely valuable. No significant gaps for a tool of this complexity.

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 single param (input.text) is fully documented in the schema. The description restates the text param ('Input: {text: string}') and lists return fields (scrubbed, redacted, kinds) with an example. It adds little beyond schema but does clarify the exact redaction behavior. 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 clearly states a specific verb+resource: 'Redact sensitive data from text before logging it.' It names exact data types (card numbers, SSNs, API keys, JWTs, PEM keys) and output markers ([REDACTED-*]). It distinguishes from AI-based scrubbing by explicitly noting 'Pure pattern matching, no AI call.'

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?

It provides clear when-to-use guidance ('Use whenever user-supplied or scraped text may carry credentials') and notes the scoping constraint ('before logging it or sending it to a model'). However, it lacks explicit when-NOT-to-use guidance or named alternatives from the sibling set, relying on implicit context.

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