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

Annotations only declare readOnlyHint=false and destructiveHint=false; the description adds the material facts an agent needs — per-frame credit cost, the free dry-run escape hatch, and that rerender is free. It also surfaces a required user-facing disclaimer about imperfect face detection. It does not say whether the source clip is overwritten or a new asset is produced, which is a notable gap for a mutating job tool.

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?

Front-loaded with the core action and the live-preset constraint, and every sentence carries distinct operational information. The dry-run, rerender, disclaimer and polling instructions are run together in a dense block without visual separation, which costs a point but nothing is wasted.

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 7-parameter mutating job tool with no output schema, the description covers the full lifecycle (start, dry-run pricing, free rerender, poll to done|failed) and the compliance message. It leaves unstated what the call returns — notably the job_id that rerender_from later depends on — and the effekt style options.

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 71%, and the description adds real meaning on top of it: dry_run returns only the price, rerender_from must be a finished job_id and makes the run free, and ausnahmen are object ids that stay sharp and only apply on rerender. The one blind spot is the effekt enum (mosaik/weich/schwarz), whose three values are left entirely unexplained in both description and 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 specific verb and resource ('Run a video VFX preset on a whole clip') and immediately narrows scope by naming the single live preset 'gesichter_verpixeln' with a plain-language gloss. It also names the polling sibling lab_vfx_preset_status, so an agent can distinguish start-vs-status without opening either 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 explicit operating order: call with dry_run=true first to see the price, and use rerender_from with a finished job_id for a free re-render with other settings. It also tells the agent to poll lab_vfx_preset_status until done|failed. It stops short of contrasting with the sibling lab_vfx / lab_vfx_status tools, so no explicit when-not guidance.

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