get_video_frames

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

With no annotations provided, the description carries full responsibility for behavioral disclosure. It does so thoroughly: it is read-only, it downloads the video and cuts frames locally, it depends on PyAV 1.7.1+ or ffmpeg, every frame becomes one model image, and frames above 12 trigger a confirmation plan rather than images. It also exposes the hard cap and token estimates, giving the agent a strong model of cost and side effects.

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 long but every sentence earns its place: it front-loads the core purpose, then explains the workflow, limits, alternative format, and each argument in a scannable Args block. Technical details like the PyAV dependency and token estimates are not padding; they affect invocation decisions. The structure mirrors how an agent actually needs to reason about the call.

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?

Given 8 parameters, no output schema, and no annotations, this description is remarkably complete. It covers what the tool returns (JPEG image blocks in time order, or JSON), when confirmation is required, the hard cap, the sheet alternative, and the sibling tool for timestamped base64 JSON. An agent has everything needed to select this tool and invoke it correctly in the tricky >12-frames case.

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 0%, so the description must compensate entirely — and it does. Every parameter gets an explanation beyond its schema title: asset_id is the video asset's UUID, count is evenly spaced frames, size includes pixel dimensions and approximate token cost, start/end are seconds, interval means one frame every N seconds, confirm is required past 12 frames, and sheet packs frames into contact sheets with timestamps. This is exemplary compensation for a schema with no parameter descriptions.

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 opens with a specific verb and resource: "Get frames of a video as image blocks, to 'watch' a clip." It clearly distinguishes itself from the sibling by explicitly naming get_video_frames_json as the alternative for base64 JSON with timestamps. This is far more than a tautology and fully differentiates the tool's scope.

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?

Provides an explicit workflow: try 6 frames first, narrow with start/end, or use interval down to 1 second. It also gives a decision rule for >12 frames (show the JSON plan, call again with confirm=true only if the user agrees), states the hard cap of 120, and names the alternative tool for a different return format. This is complete, actionable usage guidance.

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