Review a rendered video

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

Annotations cover only the safety profile (readOnly, non-destructive, closed-world), and the description goes well beyond them: external dependency requirements and the UNSUPPORTED failure mode, the fact that no Figma call is made, where images are persisted (beside the video or `saveDir`), the return ordering (JSON first, then up to maxImages images), and the semantics of declared cuts landing early/late.

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?

Dense but front-loaded: the capability statement, the required-path instruction, and the dependency caveat come before the finer rules. Few words are wasted, though the single long paragraph of interleaved clauses (holds, expectFrames, return shape, skill pointer) packs enough in that a reader may need to re-read it.

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 13-parameter, nested-object tool with no output schema, the description supplies the missing return contract (JSON numbers then findings with frames, followed by images), the dependency/environment constraints, persistence behavior, and a cross-reference to the alternative tool when dependencies are absent. Nothing an agent needs to invoke it correctly appears to be 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%, so the schema already documents most parameters, but the description adds interpretation the schema cannot: plot/frame mapping (frame k = k / fps), what `cuts` and `holds` mean for whether a result is a finding, and that `expectFrames` validates length. That is meaning beyond field definitions, though most per-field detail still lives in the schema.

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 concrete verb and resource ('See a rendered video without rendering again') and enumerates the actual outputs (contact sheet, close-ups, stills, metrics) as well as the findings it detects (freezes, flickers, flashes, jumps, cuts). It explicitly separates itself from figma_export_video (the source of `path`) and figma_check_motion (keyframe-only alternative), so an agent can pick the right sibling 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 ('Pass the `path` figma_export_video wrote'), a when-not/alternative ('Needs ffmpeg and ffprobe on PATH, else UNSUPPORTED — figma_check_motion judges keyframes without them'), and the operating condition ('answers while the plugin is busy, in any editor'). It also tells the caller to declare `cuts`/`holds` so intended jumps are not flagged, which is real usage guidance.

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