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

Annotations cover the safety profile (non-destructive, closed-world), and the description adds real substance beyond them: hard project scoping, that the project identity is echoed with the result, and an enumeration of what the audit computes (lateness, sound_off_gaps, heard_unshown, overlap/density/wrap violations). It does not clarify why readOnlyHint=false / idempotentHint=false on an audit-shaped call, leaving ambiguity about whether the compile step persists anything.

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 scope constraint is correctly front-loaded and there is no filler, but the middle section is a dense run-on of enumerated checks. Given there is no output schema, that enumeration largely earns its place as a substitute for a return-value contract.

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?

With no output schema, the description does the work of describing the returned findings, and it covers scope, ordering relative to render_preview, and purpose. The remaining gap is that limit/offset and any pagination of the report are never addressed, which matters for a tool that can emit up to 16 output times plus event pages.

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 only 33%: project_id is documented in the schema, and the description merely restates its scope constraint rather than adding syntax or semantics. limit and offset are undocumented in both the schema and the description, so paging behavior is left entirely unspecified.

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 precise verb+resource pair ('compile and audit the CURRENT caption track using the exact ASS artifact ffmpeg burns') and names concrete outputs. It is easily separable from siblings like check_speech, audit_audio_mix, find_burned_text and audit_motion without opening any schema.

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 trigger ('Call after adding or restyling captions and after render_preview') and positions itself against an alternative evidence source ('stronger timing evidence than a visual critic'). It stops short of naming the visual-critic tool by name or stating when not to use it, so it is not fully 5-level routing.

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