recognize_image_text

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

With no annotations, the description carries the full burden and does well: it discloses that OCR runs on-device, that the return includes text and normalized bounds, and that the confidence field exists only on macOS. It omits what happens when no text is detected or when the path is invalid, and says nothing about file format constraints beyond PNG.

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?

Two short sentences that front-load the operation and its mechanism before the caveat. The closing axiom about evidence vs. storage assertions is slightly jargon-flavored but earns its place as an interpretation guardrail.

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 one-parameter tool, the description covers the method, the return shape (text plus normalized bounds, macOS-only confidence), and a caution about interpretation, which compensates for the missing output schema. Only error/empty-result behavior and path conventions remain unaddressed.

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% for the single image_path parameter, so the description must compensate. It adds useful constraints ('saved simulator PNG', implying a previously captured local file), but does not clarify path format, relative vs. absolute paths, or whether the file must already exist.

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 ('Recognize text in a saved simulator PNG') plus the method ('on-device OCR'), which clearly separates it from siblings like get_screenshot or inspect_ui. It never names an alternative explicitly, so differentiation is implicit rather than spelled out.

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 'saved simulator PNG' implies you must produce a screenshot first, and 'OCR is evidence, not a replacement for storage assertions' cautions against over-relying on results. There is no explicit when-to-use vs. when-to-use-alternative guidance (e.g., versus inspect_ui or get_app_logs).

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