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

Adds substantially beyond the annotations: 1 credit per image, no charge when a 'from' is not found, refusal of official documents, 20-line cap, 50-second wait and async fallback. The unbilled-failure and refusal policies are exactly the behavioral facts annotations can't express.

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-loads purpose, then payload format, then limits, then billing/failure semantics, then timeout fallback. Every sentence carries an operational fact; nothing is filler.

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?

Covers cost, limits, refusal policy, error content and timeout behavior for a mutation tool with no output schema. Only the shape of a successful response (what IDs or result fields come back) is left unspecified, which is minor since get_edit is named for follow-up.

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 83%, so baseline is 3, but the description adds real meaning: 'from' is the text as it literally appears, 'to' empty means delete, and the 20-line limit is restated as an operational constraint. It doesn't explain hint, locale, or the image_url/image_base64 choice, leaving some parameters to 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 ('Change or remove text in a photo or screenshot') plus the quality constraint (same font, colour, background), which immediately separates it from scan_text (read-only) and get_edit (status check). An agent can route on this sentence alone.

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 preconditions and exclusions: documents like IDs, certificates, receipts are refused, and if the edit is still running 'call get_edit'. It also explains the scan-first validation path. It does not explicitly contrast with scan_text, so it falls just short of full when/when-not coverage.

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