observe_window

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

Adds substantial context beyond the readOnly/idempotent/destructive annotations: no input lock, no indicator, labels are untrusted data, refs last 60 seconds, and screenshot costs ~1-1.5k tokens. It even discloses the return shape (in_focused_input on caret line, offscreen placeholders). Only slight gap is that nothing is said about pagination/truncation of very large label sets.

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?

Front-loaded with purpose and the key distinction from run_windows, and each sentence carries distinct information (filters, item flags, screenshot cost, safety note, ref affordance). The final run-on about calling run_windows with {goal, ref} is dense and could be trimmed, but nothing is pure 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 carries the full burden and delivers: it explains the label model (UIA vs OCR), the special item flags (in_focused_input, offscreen with placeholder rect), token cost, and the ref-based handoff to run_windows. An agent has everything needed to call it and interpret results.

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 only 33%, so the description must compensate, and it does: 'contains keeps only labels containing that text (spaces ignored)' and 'screenshot true adds the window image (about 1-1.5k tokens)'. This meaning goes beyond the bare schema types for two of the three params, though target_id's schema description already covers the third.

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+resource ('Read a window's current labels') and immediately scopes it against the sibling run_windows ('as run_windows would observe them... without any input, input lock or indicator'). An agent can distinguish this read-only observation from an active run without opening any 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?

Explicitly says when to use it ('to check a result instead of starting another run') and gives conditional guidance for the screenshot param ('read the text first and ask for the image only when the text does not explain the state'). It also routes to the alternative tool with concrete instructions for acting on an item.

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