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

Annotations already declare readOnly/idempotent/non-destructive, but the description goes well beyond them: it discloses both return channels (similarity score + changed-pixel count as text, plus baseline/live/diff images) and, crucially, states the tool does NOT make the regression call and that the agent must judge. It even enumerates real regressions vs benign noise, which is exactly the context an agent needs before deciding how to act.

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 the core action and the return shape, and every clause is substantively useful. It is dense — a long single paragraph with heavy parentheticals and em-dashes — so it sacrifices some scannability, 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 carries the return-value burden and does so fully: it names the score, the changed-pixel count, and the three images returned. Combined with the input-mode explanation and the explicit 'you judge' framing, 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 100%, so the parameter baseline is 3. The description earns above baseline by clarifying the either/or relationship between baseline_image_base64 and review_id and giving the rationale ('instead of re-fetching it'), which reinforces how the two optional paths interact with the required live_image_base64.

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 (compare) and resource (two screenshots: baseline/expected vs live/current) plus the mechanism (perceptual pixel-diff). It distinguishes itself from siblings by naming list_visual_reviews as the source of a stored baseline, so an agent can place it in the visual-review workflow 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?

Clearly frames the comparison context and spells out the two input modes (two raw base64 images vs a review_id from list_visual_reviews). It stops short of explicitly contrasting against adjacent siblings like get_visual_review or resolve_visual_review, so the when-vs-alternatives guidance is implied rather than stated.

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