browser_evaluate

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

With no annotations, the description carries the full burden and it does disclose real behavior: it runs without user activation, awaits a returned Promise, and reports capture loss via capture_truncated. It does not say whether the expression can mutate page state or whether the tool is read-only, which is the most important unknown for an arbitrary-code tool.

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 core execution contract is front-loaded in the first sentence, and the following sentences each add a distinct fact (IIFE pattern, truncation signals, cursor reuse). It is dense but no sentence is pure filler; the jargon terms are only lightly defined.

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 9-parameter tool with no annotations and no output schema, the description helpfully names three return-side fields (capture_truncated, required_chars, next_item) that would otherwise be invisible. But it omits dialog_policy behavior, result_type=element semantics, and frame targeting that the schema only partly documents.

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 coverage is 56% across 9 parameters. The description adds meaning for expression, cursor (cache read, no re-execution) and the page-budget mechanic that governs max_chars, but leaves tab_handle, frame/frame_id, timeout_ms, result_type, and dialog_policy unaddressed. Partial compensation, not full.

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 opening clause 'Evaluate expression once' names a specific verb (evaluate) and resource (a page expression), which no sibling tool duplicates. It also states the execution contract (call a returned function once, await its Promise). It stops short of naming when to prefer it over browser_snapshot or browser_read_text, so it is clear but not fully differentiated.

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?

'Use an async IIFE for statements' and 'cursor reads cache without re-execution' give concrete usage hints for two scenarios. However, there is no guidance on when to reach for this tool instead of the sibling read tools (browser_read_text, browser_snapshot), and no statement of when not to use it.

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