jerrod/agent-plugins
4 skills0 installsGitHub
Install
npx skills add https://github.com/jerrod/agent-pluginsSkills in this repo
1Axraxr (Agent eXecution Readiness) is an agent skill that scores how prepared a codebase and workflow are for autonomous agent execution. It applies a 100-point rubric across twelve dimensions, combining deterministic bash-based checkers with judgment subagents so you get both objective signals and qualitative gaps. Solo builders benefit when they are about to hand a large task to Claude Code, Cursor, or similar tools and want a single readiness number instead of ad-hoc gut checks. It is not a linter replacement; it frames whether documentation, scripts, constraints, and repo hygiene support reliable agent runs. Run it during Ship review after substantial Build work, or when revisiting Operate iterate loops before another agent marathon.0installs2RevueRevue is a Claude Code development plugin that spins up a four-agent review team—architect, security, correctness, and style—working in parallel on your diffs. Solo and indie builders often skip formal review because they lack a second human; Revue approximates a staff review room when you enable Claude’s experimental agent teams. Use it before merging risky changes, touching auth or payments, or shipping a public API where style and security omissions are expensive. The workflow is intentionally heavier than a quick lint skill: you trade setup (including CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1) for structured, multi-perspective feedback. It fits the ship phase first but remains useful while iterating features in build. Confirm team mode permissions and what each sub-agent can access using the Security Audits panel on this page.0installs3SdlcSDLC is a jerrod/agent-plugins development workflow that hard-forks arqu-co/rq v1.29.8 into an end-to-end delivery system for solo and indie builders. Instead of treating brainstorm, planning, implementation, review, and release as separate chat topics, it chains them with executable quality gates—each gate is a script that must pass before the agent moves on. That design matters when you are one person wearing PM, engineer, and release hat: the skill reduces hand-wavy “done” states that blow up at deploy time. Use it when you want a repeatable ritual from idea refinement through shipped artifact, especially on SaaS, APIs, or CLIs where tests and review cannot be optional. It pairs naturally with upstream brainstorming and planning skills in the rq lineage. Inspect gate scripts and what they execute on your machine via the Security Audits panel before running on production branches.0installs4TherapistTherapist is a Claude Code workflow plugin that treats stubborn agent misbehavior like a repeatable pattern rather than a random glitch. It packages a cognitive-behavioral style framework with eleven tools, ambient hooks, and a /therapist command so you can intervene when sessions keep making the same errors, skipping guardrails, or losing thread across turns. Solo and indie builders who ship with agents hit these loops during build, review, and ops work alike; this skill is for those moments when restarting the chat or yelling in ALL CAPS stops helping. Use it when you want structured diagnosis and correction wired into the session instead of one-off reminders. It complements task skills and MCP integrations by improving how the agent executes everything else. Review hook behavior and tool permissions on the Security Audits panel before trusting it on sensitive repos.0installs