superpowers

name: superpowers tier: companion description: "Law activator for the 7 Laws of AI Agent Discipline. Unified four-source dispatcher — routes tasks to the correct Law-aligned specialist across the CI plugin (tdd-workflow, verification-loop, gateguard, ralph, deploy-receipt) and four registered upstream companions (Obra superpowers, addy agent-skills, ruflo-swarm, oh-my-claudecode) so the right discipline fires automatically instead of the agent skipping a step. Product-management coverage comes from phuryn/pm-skills via an out-of-band marketplace install (see docs/THIRD_PARTY.md). Not a peer skill — a dispatcher for the others." origin: https://github.com/obra/superpowers

Superpowers — Mandatory Agent Workflows (Four-Source Dispatcher)

Superpowers enforces a structured development workflow. Skills activate automatically when their trigger conditions are met. This is not optional guidance — it is mandatory workflow.

Philosophy

AI agents skip steps, guess, and declare "done" without verifying. Superpowers blocks this by making workflow stages explicit and enforced.

What changed in v3.8.0

The dispatcher now routes across four registered marketplaces instead of relying on Obra installed separately. All four are installable from one marketplace entry:

/plugin install superpowers@continuous-improvement       # Obra's 14 workflow skills
/plugin install agent-skills@continuous-improvement      # Addy's 21 SDLC skills
/plugin install ruflo-swarm@continuous-improvement       # Agent swarm + Monitor stream
/plugin install oh-my-claudecode@continuous-improvement  # 38 skills + 19 agents

The CI plugin (this dispatcher + tdd-workflow, verification-loop, gateguard, ralph, deploy-receipt, etc.) installs by default. The four companions are opt-in — install only what you need.

Product-management coverage (PRD, OKRs, personas, GTM, etc.) is provided by phuryn/pm-skills via an out-of-band Claude Code marketplace install:

claude plugin marketplace add phuryn/pm-skills
claude plugin install pm-toolkit@pm-skills
claude plugin install pm-product-strategy@pm-skills
claude plugin install pm-product-discovery@pm-skills
claude plugin install pm-market-research@pm-skills
claude plugin install pm-data-analytics@pm-skills
claude plugin install pm-marketing-growth@pm-skills
claude plugin install pm-go-to-market@pm-skills
claude plugin install pm-execution@pm-skills

See docs/THIRD_PARTY.md for plugin-by-plugin scope.

The Basic Workflow

The agent checks for relevant skills before any task. These are mandatory workflows, not suggestions.

Four-Source Routing Table

When a task trigger fires, the dispatcher resolves to the first available skill in the preference chain. Order = preference order. Items prefixed with ci: are bundled in this plugin; others are namespaced by their installed plugin name.

When no installed plugin in the chain resolves, the dispatcher falls back to the inline protocols below (Test-Driven Development, Brainstorming, Plan Format, etc.) so the workflow still works on a clean install.

Companion-Preference Override

The four-source routing table above is CI-first by default: where a CI-bundled skill (ci:tdd-workflow, ci:verification-loop, ci:planning-with-files, ci:context-budget, ci:ralph, ci:learn-eval) and a companion skill resolve the same trigger, the table lists the CI skill first and the dispatcher picks it. That is the conservative default — the CI plugin ships with the marketplace, the companions are opt-in installs, so a clean install routes to skills that are guaranteed present.

Once the operator has explicitly installed a companion plugin (superpowers@continuous-improvement, agent-skills@continuous-improvement, ruflo-swarm@continuous-improvement, oh-my-claudecode@continuous-improvement), the default starts working against them: the installed companion is shadowed by the CI fallback for the same trigger. This override flag respects the operator's explicit install choice without forcing every user to relitigate routing per task.

Setting the flag

Add a continuous_improvement.companion_preference key to ~/.claude/settings.json:

{
  "continuous_improvement": {
    "companion_preference": "companions-first"
  }
}

Valid values:

Which rows the override affects

These are the routing rows where the override changes the resolved target. Rows not listed here are CI-only or companion-only and route the same under any setting.

superpowers:writing-plans already wins the planning chain under both settings — it is the first entry, with ci:planning-with-files as the third fallback — so that row is unchanged.

Hard halts that remain regardless of the flag

The override does not disable:

• gateguard PreToolUse fact-list enforcement (Law 1, runtime layer in hooks/gateguard.mjs)
• Stop-hook three-section-close discipline (Law 4 + Law 7, runtime layer in hooks/three-section-close.mjs)
• Dispatcher commitments 1–6 above (subagent-driven-development default, parallel fan-out routing, TDD RED-GREEN-REFACTOR, worktree isolation, finishing-a-development-branch before push, distinct Obra/CI variants)
• Phase 0 P-MAG in proceed-with-the-recommendation (Law 5 + Law 7)

These are framework invariants, not routing preferences. The flag re-orders which specialist runs; it does not weaken what the framework guarantees.

Runtime enforcement

The override is enforced at the PreToolUse layer by hooks/companion-preference.mjs. On every Skill tool call, the hook reads ~/.claude/settings.json for the companion_preference value and:

• ci-first (default): no-op, allow.
• companions-first: emit a one-line stderr advisory naming the preferred companion; allow.
• strict-companions: block the call. Reason names the companion when its plugin is installed, or the /plugin install <plugin>@continuous-improvement hint when it is not.

The override map inside hooks/companion-preference.mjs stays row-aligned with the "Which rows the override affects" table above. Drift surfaces in the hook test suite, which walks the same pairs and fails on any new CI→companion row that the hook does not recognize.

The hook fails open. If ~/.claude/settings.json is missing, malformed, or unreadable, the hook emits { "decision": "allow" } and exits 0. Bugs in the hook never block tool calls — they only fail to enforce.

Telemetry

Every hook invocation that touches a mapped CI skill appends one JSONL line to ~/.claude/instincts/<project-hash>/companion-preference.jsonl. The line carries ts, mode, action, ci_skill, companion, plugin, and companion_installed. The action enum is:

• observation — mode is ci-first; this is the shadow row showing what companions-first would have done.
• advisory — mode is companions-first; stderr advisory was emitted.
• block — mode is strict-companions; companion installed; tool call blocked.
• block-not-installed — mode is strict-companions; companion missing; blocked with install hint.

Non-mapped skills and non-Skill tool calls write nothing. The writer wraps appendFileSync in try/catch — telemetry failure never changes the hook decision, preserving the fail-open invariant. The JSONL file is the evidence base for a future default-flip decision: after a 7-day window, the operator can grep / aggregate the file to see which routing rows the override fires on, how often, and whether the companion was installed at the time.

Stacked-PR Plan Precondition (≥3 files)

Any change touching three or more files — across skills/, src/, bin/, commands/, or any combination — must produce a stacked-PR plan as a precondition to the first edit landing. The 28-day usage report shows a clean correlation: sessions that opened with a stacked-PR plan landed at fully_achieved; sessions that began as a single big-bang multi-file edit landed at partially_achieved (landing-page dark theme, market-data-hub wiring, RAG misrouting). Single-concern PRs are the lever that closes that gap.

The required plan output has four components, in this order:

1. Per-PR table — title, scope (files), test strategy, merge order. One row per PR.
2. Dependency graph — which PRs depend on which (or "independent" if none).
3. Worktree per PR — branch name + base commit. Sequential by default; parallel only when items share no state.
4. Out-of-scope list — anything explicitly NOT in the train. Drive-by temptations get logged here, not implemented.

The plan ships as the FIRST commit of the train's first PR (under docs/plans/YYYY-MM-DD-<slug>.md) and is cited by every subsequent commit it produces.

When this rule does NOT fire

The threshold targets multi-concern feature work, not high-volume mechanical changes. The rule does NOT fire on:

• Markdown-only commits — README, CHANGELOG, docs/ updates that touch many files but ship one concern.
• Lockfile-only commits — package-lock.json, pnpm-lock.yaml, Cargo.lock, etc. updated in isolation by a dependency bump.
• Generated-only commits — output of npm run build, tsc, codemod sweeps, or any tool whose input is one source file and whose output is many derived files. The source change is what counts toward the threshold; the regenerated artifacts ride free.
• Vendor-snapshot refreshes — third-party/<name>/ updated by a documented bin/refresh-third-party.mjs driver. The single source of change is the upstream SHA bump.
• Skill-mirror sync commits — skills/<name>.md + plugins/continuous-improvement/skills/<name>/SKILL.md count as one file pair, not two, since the CONTRIBUTING.md skill mirror rule treats them as the same concern.

If you are unsure whether the rule applies, the fall-through default is to write the plan. A 30-line plan doc is cheap; a stranded big-bang edit is expensive.

Skill Library

Testing

• test-driven-development — RED-GREEN-REFACTOR cycle (includes testing anti-patterns reference)

Debugging

• systematic-debugging — 4-phase root cause process (includes root-cause-tracing, defense-in-depth, condition-based-waiting techniques)
• verification-before-completion — Ensure it's actually fixed

Collaboration

• brainstorming — Socratic design refinement
• writing-plans — Detailed implementation plans
• executing-plans — Batch execution with checkpoints
• dispatching-parallel-agents — Concurrent subagent workflows
• requesting-code-review — Pre-review checklist
• receiving-code-review — Responding to feedback
• using-git-worktrees — Parallel development branches
• finishing-a-development-branch — Merge/PR decision workflow
• subagent-driven-development — Fast iteration with two-stage review (spec compliance, then code quality)

Deployment Verification

• deploy-receipt — Closes the merge-to-production gap on auto-deploy targets (CI-side companion to finishing-a-development-branch)

Meta

• writing-skills — Create new skills following best practices (includes testing methodology)
• using-superpowers — Introduction to the skills system

Test-Driven Development (Mandatory)

RED: Write failing test → Watch it fail
GREEN: Write minimal code → Watch it pass
REFACTOR: Improve while staying green → Commit

Code written before tests is deleted. No exceptions.

Brainstorming Protocol

When given a vague request:

1. Ask clarifying questions
2. Explore 2-3 alternative approaches
3. Present design in sections
4. Wait for explicit approval before proceeding

Plan Format

Every task in a plan must include:

- task: "Specific action"
  file: "exact/path/to/file.ext"
  code: "Complete implementation"
  verify: "How to confirm it works"
  estimate: "2-5 minutes"

Code Review Severity

Subagent Development

Two-stage review for every task:

1. Spec compliance — Does it match the plan?
2. Code quality — Is it clean, tested, and maintainable?

Both must pass. Fresh subagent per task prevents context pollution.

Git Worktrees

Each feature gets an isolated workspace:

git worktree add -b feature-name ../feature-name

Clean separation, parallel development, easy cleanup.

Using Superpowers

Superpowers skills activate when their trigger conditions are detected. The dispatcher resolves each trigger through the four-source routing table above:

• "Create a feature" → superpowers:brainstorming → superpowers:writing-plans → superpowers:executing-plans
• "Fix this bug" → superpowers:systematic-debugging → superpowers:verification-before-completion
• "Review this PR" → superpowers:requesting-code-review
• "Draft a PRD" / "Write OKRs" / "Build a persona" → install phuryn/pm-skills (out of band — see docs/THIRD_PARTY.md)
• "Run this PRD autonomously" → ci:ralph
• "Fan out parallel provider migration" → superpowers:dispatching-parallel-agents or /swarm (PR D)
• "Visual regression check the landing page" → oh-my-claudecode:visual-verdict

No manual skill selection. The framework detects the trigger, resolves the chain, and enforces.