run_browser

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

Annotations give only openWorldHint/idempotentHint/destructiveHint, but the description discloses far more: tab lifecycle (opened at start_url, closed at end unless kept), failure semantics (run stops at first goal not done, rest returned in remaining_goal_ids), allowed_origins defaulting rules, fill_values meaning, and the contents of final_state and out_of_reach. Nothing here contradicts the annotations.

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?

Purpose is front-loaded in the first sentence and each subsequent sentence carries distinct operational content (tab lifecycle, failure semantics, origin scoping, result shape). It is a dense block rather than a bulleted structure, which costs a point, but no sentence is filler.

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?

With no output schema, the description still enumerates the return payload (final_state with url/title/rejected fields/downloads/form values/text lines, unmet_effects, remaining_goal_ids, tab_id, out_of_reach/screen_target). For a 12-parameter, nested-object, open-world tool this is unusually complete.

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 only 8% across 12 parameters, so the description must compensate. It explains goals ordering, keep_tab, tab_id continuation, start_url omission, allowed_origins defaults, and fill_values — a solid share — but says nothing about run_id, deadline_ms, action_budget, provider_attempt_budget, or denied_operations. Useful but incomplete compensation 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?

Opens with a specific verb+resource+scope ('Run one or more browser goals in order') and immediately distinguishes itself from siblings by naming observe_browser ('a separate observe_browser is needed only for more than that') and run_windows ('screen_target, the window run_windows can act on'). An agent can route between these tools without opening a schema.

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?

Explicit when-to-use and when-not-to-use guidance: put every known step, including post-click navigation, in one run; a later goal never runs after an earlier failure; use a separate observe_browser only when final_state is insufficient. It also states the tab continuation contract (keep_tab true → tab_id → later run with that tab_id, start_url may be omitted).

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