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

Annotations already establish readOnlyHint=true, idempotentHint=true, destructiveHint=false, and openWorldHint=false, so the safety profile is covered. The description adds real value by naming the threat classes it catches and the endpoint, but stays silent on the most operationally important behavior: what happens on a violation (block, redact, annotate) and what the caller receives back. The word "sanitize" sits in mild tension with a read-only hint, though it is reasonably read as a stateless transform rather than a stored mutation.

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?

One front-loaded sentence states the capability and threat coverage, with the endpoint appended compactly. Nothing is padded, though the POST path is largely redundant for an MCP caller and the parenthetical could have been spent on return behavior instead.

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 3-parameter, read-only guard with no output schema, the description should explain the result shape (verdict flags vs sanitized text) and the semantics of the two list parameters. It covers what is scanned but not what comes back or how findings are reported, leaving a meaningful gap for an agent deciding whether to gate on the response.

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 nominally 100%, which sets the baseline at 3, and the description does not mention any parameter. This is unfortunate because the schema's own descriptions are placeholders ("Array parameter 'canary_tokens'", "Array parameter 'allowed_link_domains'"), so the meaning of canary tokens and the domain allowlist is documented nowhere—neither schema nor description explains that canary tokens are tripwires whose reappearance signals leakage.

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 names a specific verb pair ("Scan and sanitize") and a specific resource ("LLM outputs"), and enumerates the hazards checked: prompt leakage, Markdown beacons, XSS, and policy violations. It is clearly distinct in substance from agent-side or prompt-side siblings, though it never explicitly contrasts itself with tinplate_prompt_guard or tinplate_llm_messages_guard.

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?

Usage is only implied: an agent can infer this runs over model output before it is shown or forwarded. There is no explicit when-to-use statement, no exclusions, and no routing guidance toward alternative guard tools, so the agent must infer placement in a pipeline.

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