library/superpowers
Superpowers Extended Methodology (Library) reference
The Superpowers Extended methodology provides a complete agentic software development lifecycle with structured skills, subagent-driven execution, and quality gates at every stage. It enforces disciplined workflows: brainstorm before coding, test before implementing, verify before claiming completion.
Continue reading
Nearby pages in the same section.
Documented graph nodes
Records linked directly from this page’s Page node.
Superpowers Extended Methodology
**Source**: pcvelz/superpowers (Community fork of obra/superpowers) **Category**: Full Development Lifecycle / Agentic Software Development **License**: MIT
Overview
The Superpowers Extended methodology provides a complete agentic software development lifecycle with structured skills, subagent-driven execution, and quality gates at every stage. It enforces disciplined workflows: brainstorm before coding, test before implementing, verify before claiming completion.
Process Files
| Process | File | Description | Task Count |
|---|---|---|---|
| Full Workflow | superpowers-workflow.js | Complete lifecycle orchestrator | 12 |
| Brainstorming | brainstorming.js | Socratic design refinement | 5 |
| Writing Plans | writing-plans.js | Bite-sized TDD task planning | 4 |
| Executing Plans | executing-plans.js | Batch execution with checkpoints | 6 |
| Subagent-Driven Dev | subagent-driven-development.js | Fresh agent per task + two-stage review | 6 |
| Test-Driven Dev | test-driven-development.js | RED-GREEN-REFACTOR cycles | 7 |
| Systematic Debugging | systematic-debugging.js | 4-phase root cause process | 9 |
| Parallel Agents | dispatching-parallel-agents.js | Concurrent domain solving | 4 |
| Verification | verification-before-completion.js | Evidence-based completion | 3 |
Skills Catalog
| Skill | Directory | Description |
|---|---|---|
| brainstorming | skills/brainstorming/ | Design before code, Socratic refinement |
| writing-plans | skills/writing-plans/ | Bite-sized TDD implementation plans |
| executing-plans | skills/executing-plans/ | Batch execution with checkpoints |
| subagent-driven-development | skills/subagent-driven-development/ | Fresh agent per task, two-stage review |
| test-driven-development | skills/test-driven-development/ | RED-GREEN-REFACTOR iron law |
| systematic-debugging | skills/systematic-debugging/ | 4-phase root cause investigation |
| dispatching-parallel-agents | skills/dispatching-parallel-agents/ | One agent per independent domain |
| using-git-worktrees | skills/using-git-worktrees/ | Workspace isolation |
| finishing-a-development-branch | skills/finishing-a-development-branch/ | Branch completion workflow |
| requesting-code-review | skills/requesting-code-review/ | Dispatch code reviewer |
| receiving-code-review | skills/receiving-code-review/ | Handle review feedback |
| verification-before-completion | skills/verification-before-completion/ | Evidence before claims |
| writing-skills | skills/writing-skills/ | TDD for skill creation |
| using-superpowers | skills/using-superpowers/ | Skill discovery and invocation |
Agents Catalog
| Agent | Directory | Role |
|---|---|---|
| code-reviewer | agents/code-reviewer/ | Senior code reviewer (plan alignment + quality) |
| implementer | agents/implementer/ | Task implementation subagent (TDD, self-review) |
| spec-reviewer | agents/spec-reviewer/ | Spec compliance verification |
| code-quality-reviewer | agents/code-quality-reviewer/ | Code quality assessment (after spec passes) |
Workflow Lifecycle
brainstorming -> writing-plans -> [setup worktree] -> executing-plans OR subagent-driven-development -> verification -> finishing-a-development-branchCross-cutting concerns applied throughout:
test-driven-development- All implementation follows TDDsystematic-debugging- All bug investigation follows 4-phase processverification-before-completion- All claims require evidencerequesting-code-review/receiving-code-review- Review between tasks
Philosophy
- **Test-Driven Development** - Write tests first, always
- **Systematic over ad-hoc** - Process over guessing
- **Complexity reduction** - Simplicity as primary goal
- **Evidence over claims** - Verify before declaring success
- **YAGNI ruthlessly** - Remove unnecessary features from all designs