labeling_tasks

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

No annotations are provided, so the description carries the full burden. It discloses the manifest's content shape, which is genuinely useful, but says nothing about whether a file is written, where, whether it overwrites existing output, or whether the operation is read-only or mutating.

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?

A single front-loaded sentence: the action first, then the manifest's structure, then the consumer. Every clause adds information and nothing is padded.

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 five parameters, 20% schema coverage, no annotations, and no output schema, the description leaves too much unsaid — notably what 'z', 'max_tasks', and 'include_segments' control and whether a file is produced at 'out'. Adequate for understanding the artifact, inadequate for calling it correctly.

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 20%, so the description must compensate, and it does not: 'path', 'z', 'max_tasks', and 'include_segments' are never mentioned. Only 'out' has a schema description, so most parameters are opaque.

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 artifact ('Generate an annotation manifest') and even previews what each task contains (episode, frame range, cues, question). That is far more specific than any sibling in the list, though it never explicitly contrasts itself with tools like segment_episodes or plan_windows.

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?

The phrase 'ready for a human reviewer or a vision-language model' hints at the downstream consumer but gives no when-to-use condition, no prerequisites (e.g. must the dataset be detected/segmented first?), and no alternatives. An agent must infer placement in the pipeline.

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