Prose Coach

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

The description discloses determinism ('the same text always returns the same findings'), which is valuable behavioral transparency beyond what annotations provide. The readOnlyHint=true annotation confirms no side effects, and the description is consistent with that. It also discloses the return structure (each finding with evidence quoted and the fix), adding context not present in annotations. It doesn't mention auth or rate limits, but for a deterministic read-only scanner these aren't critical gaps.

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 covering purpose, deterministic behavior, usage timing, and return format. Dense but genuinely informative with zero filler. Could arguably be split, but everything earns its place. The enumeration of pattern types is useful, though the return-format sentence could be tightened. No fluff or redundancy.

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, deterministic, read-only scanner with a full description and good annotations, this is well covered. The description explains scope, behavior, usage timing, and output format. With no output schema present, the description's mention of return structure (evidence quoted, fix) helps compensate. Minor gaps: doesn't specify what happens on empty or very long inputs, but these are low-stakes for this tool type.

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% (only one 'text' parameter with a clear description). The description adds guidance over the schema: it tells the user to paste 'the prose itself, not a summary of it', which adds real meaning beyond the schema's 'The draft to scan' phrasing. Since the description and schema already carry meaningful detail for a single-parameter tool, the marginal value added is solid.

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 what the tool does: 'Scan a draft for the patterns that make writing read as AI-generated' with specific enumerated patterns (banned vocabulary, structural tells, pattern density). It uses a specific verb (scan) plus a concrete resource (draft) and is further enriched with the deterministic note and return format (evidence + fix). There are no siblings, so no differentiation is needed.

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 guidance is explicit: 'Run this before showing the user any outward-facing draft' with concrete examples (email, post, cover letter, marketing copy, blog article). This gives clear when-to-use direction. There are no sibling tools to contrast against, so exclusions aren't needed, but the 'before showing' timing instruction is genuinely actionable and context-rich.

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