AI thao tác trình duyệt

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

No annotations are supplied, so the description carries the full burden. It does disclose two genuinely useful behaviors beyond the schema: that clickAt, fill and press are gated behind allowWrites=true, and that only fake QA data may be used. It says nothing about what happens to page state, whether a snapshot is returned, or error behavior when a selector/role fails to match.

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 short, front-loaded sentences: capability list first, then the write-gating condition, then the data-safety rule. No filler, though the gating sentence could name the parameters it governs more precisely instead of restating the action names.

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 10-parameter mutation tool with zero annotations and no output schema, the description is thin. It never says which parameters are required, what the tool returns after an action, whether a browser session/page must exist first, or how the scroll delta behaves — all things an agent needs to call it correctly on the first attempt.

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 10 parameters, so the description must compensate and only partly does. It conveys role/selector targeting, x/y coordinates from an image, and a fill value, but the required 'project' parameter, 'page', 'name', and 'deltaY' (direction, magnitude, units for scroll) are never explained, leaving nearly half the surface undocumented.

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 concrete set of browser verbs that map almost one-to-one onto the action enum (navigate, click, clickAt, fill, press, hover, scroll), so an agent can tell what the tool does. It does not contrast itself with siblings such as tester_browser_see or tester_browser_assert, leaving the boundary between acting and observing to inference.

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?

Usage context is implied rather than stated: a third of the actions are flagged as needing allowWrites=true, which functions as a precondition, and 'Chỉ dùng dữ liệu QA giả' is a scope restriction. There is no explicit when-to-use versus when-to-use-an-alternative guidance, and no statement that a session must first be started.

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