Android tap Text

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

Annotations already mark destructiveHint=true, but the description adds crucial context: 'On a real phone every tap has real consequences (posts, likes, purchases, messages): NEVER tap an unidentified control to find out what it does.' It also explains the matching rules, the 300ms wait, the screenshot refresh, the polling behavior for expect_text/expect_gone, and the pixel-coordinate detail (no scale/rotation inverse). This goes well beyond the annotations by describing exactly what happens during execution, which is essential for a destructive action.

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?

The description is dense but well-structured. It front-loads the core action, then layers matching rules, usage guidance, cautionary notes, technical pixel details, and confirmation method. Every sentence adds value; there is no redundancy or filler. Despite its length, it remains readable and purposeful.

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?

This is a complex tool (OCR, tapping, polling, ambiguity handling) with no output schema, so the description must carry the burden of explaining behavior and return values. It does: it mentions the fresh screenshot summary shape (same as android_interact), the expected.matched report, and the candidate-list error for ambiguous matches. It also covers the safety stop condition. For an agent to call this correctly, nothing essential 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 covers 100% of parameters with descriptions, so baseline is 3. The description adds meaningful semantics for the key parameters: it explains the query matching rules (exact → case-insensitive-contains → candidate-list ambiguity), and clarifies the purpose of expect_text/expect_gone as a single-round-trip confirmation mechanism. It doesn't repeat schema details but enriches them with usage context, justifying a score above baseline.

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?

The description states exactly what the tool does: OCR the current screen and tap the center of the best text match, with specific matching rules that mirror android_tap_element. It also clearly differentiates this tool from its sibling by specifying it works for text the view hierarchy cannot see, and explicitly says to prefer android_tap_element when resource-id or content-desc exists. This is a specific verb+resource+scope that leaves no ambiguity.

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 description gives explicit guidance on when to use this tool vs android_tap_element: 'Prefer android_tap_element whenever the control HAS a resource-id or content-desc: identity beats pixels.' It also warns against using this tool on unidentified controls and tells the agent to STOP and report instead. Additionally, it explicitly says not to screenshot-and-compare pixels to verify a tap, steering the agent to use expect_text/expect_gone instead. This is comprehensive usage direction.

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