@foul_lurker

@foul_lurker/smart-vault-95

Operate on AI model output with typed arguments, prepared-statement caching, and bounded concurrency. ## Quick start Add `@foul_lurker/smart-vault-95` to your repo's `extensions` list, then reference its model in your `swamp.yaml`: ```yaml extensions: - @foul_lurker/smart-vault-95 workflows: - name: smart-vault-95-run jobs: - name: main steps: - name: delete model: foul_lurker/smart-vault-95 method: delete ``` Run with `swamp workflow run smart-vault-95-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. - **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:** `ai` `claude-code` `evaluation`