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

Annotations provide readOnlyHint=true, and the description builds on that with rich behavioral context: it reads changed-file content directly from the PR's source commit, embeds the FULL un-merged source when includeSource is true (one fenced block per file, trucated past sourceCharsPerFile), and explains why metadata-only PRs may show few BP violations. It also discloses auth needs (DEVOPS_ORG_URL + DEVOPS_PAT, Code: Read) and the side-effect-free workflow (produces a ready-to-post comment that a separate tool posts with user confirmation). Nothing contradicts the read-only annotation.

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 front-loaded with the PRIORITY TRIGGER block and every sentence carries operational directives an agent needs (triggers, exclusions, mechanism, edge case, workflow, auth). There is some redundancy — the NEVER/DO NOT git directives are re-iterated and the embedded-source benefit is stated twice — and the capitaliation-heavy style adds noise. Overall the length is justfied by the tool's complexity.

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 complex analysis tool with no output schema, the description defines what the agent gets back: BP Critical/Warning violations, upgrade impact cross-referencing CoC targets and event handlers, extension-conflict findings, embedded source, and a ready-to-post PR review comment adressed to the author. It also covers edge cases (metadata-only PRs, larger PRs summarized beyond maxFilesDeep), auth prerequisites, and the next tool to call. An agent has everything needed to select and invoke this correctly in one read.

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 100%: all 7 parameters (prId, repositoryId, project, maxFilesDeep, includeSource, includeWarnings, sourceCharsPerFile) carry their own schema descriptions, including defaults and behavioral notes. The tool description adds only marginal value beyond that — the 'one fenced block per file' formatting detail and a re-statement of the includeSource rationale. Since the schema already does the heavy liftng, the baseline 3 applies.

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 names a specific verb and resource ('Analyse the full D365 F&O code impact of a Pull Request') and explicitly carves out the tool's scope: it reviews proposed un-merged code including brand-new files via the ADO REST API. It further distinguishes itself from search_d365_code and local git diffs, so an agent can tell what this tool is and is not. The only minor gap is that the similarly-named sibling ado_review_xpp_pr is never adressed, but the core purpose statement is unambiguous.

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?

Guidance is exceptionally explicit: a literal trigger-phrase list ('analyse PR', 'review PR', 'PR #', 'D365 impact of PR'), a hard exclusion ('NEVER call search_d365_code when PR or Pull Request + a number is mentioned'), and an explicit prohibition on git fallbacks ('DO NOT fall back to local git show/git diff', 'NEVER run git'). It also routes the follow-up to ado_post_pr_comment with the user-confirmation requirement. The one omission is not routing to the sibling ado_review_xpp_pr, but the trigger list effectively decides selection for PR-review intents.

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