ui2.screen_evidence

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 declares that the tool only produces evidence and never asserts state, and it discloses the concrete signals computed per category. It omits output format, cost, and permission requirements, keeping it short of a 5.

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 the purpose, then lists evidence modes compactly, then closes with the discipline rule. Dense but every clause earns its place; the colon-delimited enumeration is terse rather than bloated.

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 complex six-mode tool with no output schema and no annotations, the evidence-category explanation is solid, but the undocumented 'game' parameter and the absent return-shape description leave real gaps an agent must guess at.

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?

The single parameter 'game' has 0% schema description coverage and is never mentioned in the description, so an agent gets no meaning for it from either source. The description fails to compensate for the coverage gap.

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+resource (屏幕证据采集 / screen evidence collection) and enumerates the six evidence categories it computes (overlay, motion, color, filter, layout, hud). The clause 只出证据不下结论 ('only emits evidence, does not conclude') meaningfully separates it from conclusion-oriented siblings like ui2.state and ui2.check_states, though it never names them explicitly.

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 anti-hallucination rule (缺证据时列备选假设) guides how to interpret results, but there is no explicit when-to-use / when-not-to-use or named alternative among the many ui2.* and vision.* siblings. Usage is implied rather than spelled out.

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