Claude Code — install as plugins via the 37signals marketplace:
claude plugin marketplace add basecamp/claude-plugins
claude plugin install dev # or: security, ai, recapOther agents — install as standalone skills:
npx skills add basecamp/house-skills| Skill | Plugin | Description |
|---|---|---|
| address-pr-reviews | dev | Address PR review comments - fix issues, reply to threads, mark resolved |
| basecamp-activity | recap | Fetch Basecamp project or person activity into day-cached JSON atoms |
| consult-outside-expert | dev | Explicitly launch an independent outside-expert consultation and refinement loop |
| git-activity | recap | Fetch git log from local repos into day-cached JSON atoms |
| github-activity | recap | Fetch GitHub PRs, reviews, issues, and commits into day-cached JSON atoms |
| ralph-lisa-loop | dev | Explicitly launch the automated plan-implement-review loop with rope-length control |
| harden-github-actions | security | Resolve zizmor warnings in GitHub Actions workflows, harden CI pipelines, and pin actions to SHA hashes |
| agents-md | ai | Write, slim, or audit a repository's agent instruction file — what earns always-on context and what belongs behind disclosure |
| install-md | ai | Create install.md files optimized for AI agent execution |
| ruby-cext-memory-truffle-hunt | dev | Scent library for use-after-free and GC-invalidated pointers in Ruby C extensions |
| recap | recap | Activity digests — pluggable fetchers, timescale synthesis, audience-aware composition |
| skill-crafting | ai | Create and refine agent skills through co-development and eval loops |
| truffle-hunt | dev | Sweep a dependency corpus for every instance of one latent bug class — scoping, proof protocol, and filing upstream |
Real skill files live in plugins/{plugin}/skills/. The top-level skills/ directory
contains symlinks for a unified view. See AGENTS.md for details.