Rapid Pulse 56

Read, reconcile, and report on LLM responses across multiple environments. Idempotent and dry-runnable.

Quick start

Add @openai/rapid-pulse-56 to your repo's extensions list, then reference its model in your swamp.yaml:

extensions:
  - @openai/rapid-pulse-56

workflows:
  - name: rapid-pulse-56-run
    jobs:
      - name: main
        steps:
          - name: delete
            model: openai/rapid-pulse-56
            method: delete

Run with swamp workflow run rapid-pulse-56-run. The first invocation prints what it would change; pass --apply to commit.

What's inside

• Typed models — Zod-validated arguments, structured outputs, and a manifest the agent can reason about.
• Reports — markdown roll-ups grouped by label and breaking-change flag.
• Skills — Claude Code skills with tool whitelists and trigger-phrase manifests.

Configuration

Configure via ~/.swamp/config.yaml or environment variables. The full method reference and rubric format live in the package's manual/ directory.

Labels: llm evaluation