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

Far exceeds what the annotations cover (annotations only say non-readonly, open-world, non-idempotent, non-destructive). The description discloses async job semantics with persisted-on-disk state that survives tab closes, the return shape {jobId, fileName} with format decided up front, that ElevenLabs SPENDS the user's credits, and the full structured-refusal catalogue (NO_API_KEY, INSUFFICIENT_CREDITS, CONSENT_REQUIRED with the await_user_action retry loop, PERMISSION_REQUIRED), plus the audio-tag stripping safety net.

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?

Well front-loaded: purpose and outputs first, then the workflow rationale, then the premium/error handling. It is long, and the audio-tag rule appears in both the body and the modelId schema text, which is mild redundancy — but for an 11-param tool with a consent flow, nearly every sentence carries operational weight.

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 11 parameters, a nested object, no output schema and a premium path, the description covers everything an agent needs: async return values, how to poll (get_job_status), the mediaRef on completion, the consent handshake, provider prerequisites, and the segmentation advice that prevents a bad edit. Nothing material is 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 description coverage is 100%, including the nested voiceSettings object, so the schema already carries the heavy lifting — baseline 3 applies. The description does reinforce a few behaviors (audio tags only valid when the model reports supportsAudioTags, consentActionId is good for one job, language ignored for ElevenLabs), but these are mostly repeated from the schema rather than adding new syntactic detail.

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 ('Generates an AI voiceover from a script') and immediately distinguishes the two provider paths (on-device free models vs. cloud ElevenLabs). It also spells out the concrete side effects — WAV in the media library, script stored as metadata, word-level transcript written — so an agent knows exactly what this produces versus siblings like add_sfx or start_transcription.

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?

Explicitly says when to do this (before picture is cut, because a script doesn't reveal its spoken duration) and names concrete downstream steps (get_transcript for word timings, read_skill {topic:'voiceover'}, list_voiceover_models for voices). It also gives exclusions and preconditions: check the elevenlabs status object first, a local model must be downloaded and at phase 'ready', and v2 only when the user asks.

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