Tools provided
Tools the agent exposes — your AI client calls them automatically when it needs them.
- run_full — Orchestrated path: from the repo scan you provide, score per-file risk, split the work into a series of independently mergeable 20 to 40 file pull requests, and return the full plan with the PR series, progress board and characterization-test coverage.
- analyze — Run pure deterministic per-file migration-risk analysis over a provided repo scan with no model call and return scored findings, worst first.
- plan_pr_series — Chunk a provided repo scan into reviewable, independently mergeable 20 to 40 file pull requests, each with its worst-case risk level and its characterization-test requirement.
- equivalence_test_coverage — Return the fraction of planned pull requests that require a behaviour-equivalence test before merge, computed from the provided scan.
- list_capabilities — List the agent's static capabilities: the available tools, the supported migration campaigns and the risk-classification tiers. Useful for discovery.
- plan_inputs — Plan the inputs for a tool: returns the clarifying questions, the JSON schema and a ready-to-edit example.
- discover_intent — Understand your goal and co-design the exact input through clarifying questions before running.
- open_form — Opens the guided form to collect the repository scan, the migration type and the batch size. The entry point for running the agent on your own data.
- run_form — Internal: invoked by the guided form when the user submits. Not for direct use.
