superpowers
claudecodexCore skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
15 skills · hooks
Scope a project by interview, plan each gate on plan mode with an adversarial reviewer, slot findings in without ad-hoc edits, break every gate with a fresh red-lane agent, and hand off only what the product owner ran and accepted. Hooks are silent outside projects that opt in.
Plugin serial-build from marketplace serial-build-harness in DMG-Venture-Studio/serial-build-harness. Commands read from the repo's manifest — 12 days ago.
This plugin runs code on your machine — its hooks fire automatically on events like tool use. The exact commands are listed below; read them before installing.
demo-briefWrite the instructions the product owner follows to run a gate's demo, including the step where they try to break it, with every fact computed from the project rather than remembered. Use when a gate is built and red-checked and the owner is about to run its demo.
gatePlan the next roadmap gate on Claude's plan mode, after an adversarial review of every draft, and write docs/plan.md on the product owner's approval. Use whenever a gate comes up for planning in a serial-build project; never another planning skill for a gate.
handoffClose a session so a fresh agent picks up the work at once. Re-runs every proof in docs/done.md, refuses to move a gate while a red finding still reproduces, moves an accepted gate to done, and records state in the gate ledger. Use at the end of a working session or when the product owner has accept
intakeSlot a new finding, idea, bug, requirement, or pasted brief into a serial-build project's docs without editing the plan or the standing docs ad hoc. Use whenever something new turns up mid-work - a bug, an idea, a review's findings, a handoff from another session.
redSend a fresh agent that did not build the active gate to break it, from its own git worktree. Its only outputs are a failing test, brought back and committed, or "nothing found" with every attempt listed; it never edits the fix. Use on every gate before the handoff, after the builder believes the ga
scopeInterview the product owner and scaffold a project for the serial-build loop - vision, first demo, floors, stack, where things live, the docs ledger, rules, and the hook config. Use on a new or unscaffolded repository when the user asks to scope, set up, or start a project with serial-build.
tooling-fixFix the project's check-tooling repairs that intake queued under "## Tooling fixes" in docs/roadmap.md, in one pass of their own, from a worktree, outside any product gate. Use when intake queued tooling fixes, or the owner asks to fix the check-tooling backlog.
adversarial-reviewerUse this agent to attack a draft before the product owner approves it - a gate plan inside /serial-build:gate, or the docs an intake wrote against the source it was given. Typical triggers include the gate skill reviewing each plan draft before ExitPlanMode, a second pass over an already-folded draf
red-breakerUse this agent only from the /serial-build:red skill, to make one gate fail. It did not build the gate, trusts no claim that it works, writes only test files, and returns either reproduced failures as failing tests or "nothing found" with every attempt listed. Typical triggers include the red skill'
red-fixerUse this agent only from the /serial-build:red skill, to fix the red findings filed in docs/plan.md's Red section, from a fresh context. It is never the red-breaker and never judges whether a finding is right. It fixes each kind of bug once, with one test that enumerates the class, never edits a red
sh "${CLAUDE_PLUGIN_ROOT}/hooks/scripts/session-start.sh"
sh "${CLAUDE_PLUGIN_ROOT}/hooks/scripts/guard-generated.sh"
sh "${CLAUDE_PLUGIN_ROOT}/hooks/scripts/red-fence.sh"
sh "${CLAUDE_PLUGIN_ROOT}/hooks/scripts/red-lock.sh"
sh "${CLAUDE_PLUGIN_ROOT}/hooks/scripts/edit-checks.sh"
sh "${CLAUDE_PLUGIN_ROOT}/hooks/scripts/inbox-touch.sh"
sh "${CLAUDE_PLUGIN_ROOT}/hooks/scripts/stop-checks.sh"
sh "${CLAUDE_PLUGIN_ROOT}/hooks/scripts/inbox-stop.sh"
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
15 skills · hooks
tasteskill 2.5.1: professional company websites at the figma.com and x.ai level, with one section per screen, real photographs, real product UI, effects kept for tech and gaming pages, and a real-browser checker and finish gate.
1 skill
5 skills + 35 modes + Material Passport pipeline. Includes v3.6.7 cross-model audit gate and v3.6.8 generator-evaluator contract.
5 skills · 16 commands · 4 agents · hooks
Backend API design, GraphQL architecture, workflow orchestration with Temporal, and test-driven backend development
9 skills · 1 command · 8 agents
AI-supervised issue tracker for coding workflows
1 skill · 1 agent · hooks
PRD generation and conversion skills for the Ralph autonomous agent loop
2 skills
This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.