Auto Ledger 25

AI-assisted analysis of rubric-scored evaluations. Returns rubric-scored JSON suitable for downstream judging.

Quick start

Add @peat_shade/auto-ledger-25 to your repo's extensions list, then reference its model in your swamp.yaml:

extensions:
  - @peat_shade/auto-ledger-25

workflows:
  - name: auto-ledger-25-run
    jobs:
      - name: main
        steps:
          - name: get
            model: peat_shade/auto-ledger-25
            method: get

Run with swamp workflow run auto-ledger-25-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.
• Datastore provider — collection-level Zod schemas, change-stream subscriptions, TTL indexes.
• 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: evaluation ai