observe_browser

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

Annotations already cover readOnly/idempotent/openWorld/non-destructive, and the description adds genuinely new behavior: screenshot brings the tab to the front and costs roughly 1-1.5k tokens, contents are untrusted data, and contains filtering ignores spaces. It does not cover error cases such as a missing or closed tab, so not a full 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?

The read action and its scope are front-loaded, and every clause carries information (return payload, filter, screenshot cost, trust warning). It is a dense run-on of semicolon-joined clauses rather than clean sentences, which slightly hurts scanning.

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 no-output-schema read tool, the description covers the return payload, the filtering parameter, the screenshot trade-off (token cost plus foregrounding) and a data-trust caveat. Only the provenance/requirements of tab_id and failure behavior remain unstated.

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?

With 0% schema coverage the description carries the load and does well: contains is defined as keeping matching items and text lines with spaces ignored, and screenshot's side effect, cost and recommended ordering ('read the text first') are explained. tab_id is only implicitly identified via run_browser(keep_tab) and never marked required.

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 (read) and resource (a browser tab kept by run_browser(keep_tab)) and enumerates what is returned (offered targets with kind/label/states plus visible page text), which separates it from observe_window and get_run_journal. The phrase 'without any input' is slightly misleading given tab_id is required, keeping it below a 5.

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?

Gives a concrete usage scenario ('to see No results or a validation error before choosing the next goal'), which implies when to inspect. It never names alternatives (observe_window, list_windows) or states exclusions, so the agent must infer routing from the scenario alone.

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