Tools provided
Tools the agent exposes — your AI client calls them automatically when it needs them.
- review_diff — Review a unified diff: parse the changed files, judge each acceptance criterion against the added lines, assess whether the changed tests actually exercise the change, and decide whether it can merge.
- check_acceptance — Judge each acceptance criterion against a diff's added lines, returning the keywords that matched so the judgement is checkable.
- check_test_evidence — Assess whether the tests in a diff actually exercise what it changed, flagging tests that assert nothing or only that the call did not throw.
- write_pr — Compose the pull-request title and body from a completed review, naming the verdict, the blockers and the evidence.
- detect_breaking_change — Compare the API contract a change was reviewed against with the contract it produces, classify every delta as breaking or non-breaking, and derive the semver impact.
- score_tech_debt — Score the maintenance pressure on the files a change touches with a published, inspectable interest-rate formula, and rank the hotspots worst first.
- run_full — Review the diff and compose the pull-request draft in one call, adding the breaking-change and tech-debt checks for whichever of their inputs you supplied. With no diff, returns a needs_input prompt rather than a demo review.
- list_capabilities — List the agent's capabilities: the tools it advertises, what each one does, and the merge verdicts it can return. Useful for discovery.
- plan_inputs — Plan the inputs for a tool: returns the questions, the schema and a ready-to-edit example.
- discover_intent — Understand your goal and co-design the exact input via clarifying questions before running.
- open_form — Use this the moment the user wants to run this agent on their own input. It opens an interactive form to collect the run arguments. Do NOT check credentials, do NOT ask clarifying questions in chat, and do NOT run any analysis before calling this — this form is the entry point for running the agent.
- run_form — Internal: invoked by the open_form form when the user submits. Reads exactly the same arguments as run_full and forwards all of them. Not for direct use.
