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

Annotations only say readOnlyHint=true; the description goes well beyond that by explaining the resolved-by semantics (commitSha/prNumber/buildId), the independent-of-review-state behavior, and especially the `removed` list which surfaces otherwise-invisible baseline deletions with lastBuildId context. It also notes the feature requirement and the absence of image URLs, both useful behavioral disclosures.

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 description is long but information-dense: every clause earns its place, and the most decision-relevant contrast (vs get_build, survives in-PR accepts) is front-loaded. It trails into a multi-item return-value inventory, which makes it slightly heavy, but that inventory includes behavioral caveats (removed stories carry lastBuildId) rather than mere schema repetition.

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?

The description covers what the tool answers, how it differs from its closest sibling, what the response contains including pagination cursors, and how to get pixels elsewhere. With an output schema present, return-value detail is appropriately summarized. For a read-only changeset query, this is complete at the right level of detail.

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 baseline is 3. The description adds value by clarifying that resolution can come from any of the three identifiers and that the result is 'resolved by commitSha/prNumber/buildId', tying the parameters to the same concept. It doesn't explain precedence among them, but the schema already documents each field adequately.

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 description opens with a precise noun phrase and verb — 'cumulative this PR vs base visual changeset' — and immediately distinguishes it from get_build by the in-PR accept behavior. It names siblings and states exactly what it returns (base, counts, first page of stories, deleted stories with lastBuildId).

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 contrasts with get_build ('unlike get_build...') and even explains the carve-out for mid-PR accepted stories, so an agent knows when to choose this tool over the alternative. It also tells the caller to page with list_pr_stories and fetch pixels with get_diff / render_diff_image, which are concrete routing instructions.

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