AI ghi nhận lỗi giao diện bằng bằng chứng hình ảnh

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

There are no annotations, so the description must fully explain the tool's behaviour. It does well to disclose that it automatically captures evidence screenshots and files the finding into a per-project QA report, and that AI-generated aesthetic judgements are flagged REVIEW_REQUIRED. However, it omits whether findings are persisted, whether duplicate calls create duplicates, what the response contains, and what happens if screenshot capture fails. For an unannotated write tool, these are significant gaps.

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?

Three sentences, front-loaded with the workflow trigger and the defect categories, with no filler. The phrase "bằng bằng chứng hình ảnh" in the title is a minor repetition, but the description itself is tight.

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 nine parameters, five of them required, no annotations and no output schema, the description covers the trigger and side effects but leaves the severity model, the expected and observed fields, and the report structure unexplained. It is adequate to attempt a call, but not to use the tool correctly in ambiguous cases.

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% across nine parameters, so the description must document them itself. It indirectly touches the category enum (layout, responsive, colour, spacing, obscured buttons), the defect-versus-improvement kind, and the project scoping. It gives no explanation of severity, expected, observed, suggestion, or reproductionSteps, which are five of the nine parameters. It therefore covers less than half of the input surface.

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 a specific verb and resource: the model "ghi lỗi bố cục, responsive, màu chữ, khoảng cách, nút che khuất" (records layout, responsive, font colour, spacing, hidden-button defects) and aesthetic suggestions. It also names its prerequisite sibling tester_browser_see, so the agent can distinguish it from a pure observation tool. It is not fully differentiated from siblings such as tester_browser_layout_audit or tester_browser_accessibility_audit.

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 a clear ordering rule: use this tool after viewing an image through tester_browser_see. However, it never states when NOT to use it, nor how it relates to layout_audit, accessibility_audit or inspect, which likely overlap. The agent must infer that this is the recording step rather than the detection step.

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