obra/superpowers
An agentic skills framework & software development methodology that works.
https://skillcdn.ai/gh/obra/superpowers내 AI에 이 주소를 연결하면 이 스킬들을 쓸 수 있어요. 연결 방법 보기
- 기본 브랜치
main - 커밋
22f0b46 - 라이선스
MIT - 검색단어와 의미
저장소 둘러보기
스킬과 문서의 원문에서 단어와 의미로 찾습니다. 화면용 번역은 검색되지 않아요.
문서
- Document Review System DesignAdd two new review stages to the superpowers workflow:docs/superpowers/specs/2026-01-22-document-review-system-design.md
- Visual Brainstorming Refactor: Browser Displays, Terminal CommandsDate: 2026-02-19 Status: Approved Scope: lib/brainstorm-server/, skills/brainstorming/visual-companion.md, tests/brainstorm-server/docs/superpowers/specs/2026-02-19-visual-brainstorming-refactor-design.md
- Zero-Dependency Brainstorm ServerReplace the brainstorm companion server's vendored node_modules (express, ws, chokidar — 714 tracked files) with a single zero-dependency server.js using only Node.js built-ins.docs/superpowers/specs/2026-03-11-zero-dep-brainstorm-server-design.md
- Codex App Compatibility: Worktree and Finishing Skill AdaptationMake superpowers skills work in the Codex App's sandboxed worktree environment without breaking existing Claude Code or Codex CLI behavior.docs/superpowers/specs/2026-03-23-codex-app-compatibility-design.md
- Worktree Rototill: Detect-and-DeferDate: 2026-04-06 Status: Draft Ticket: PRI-974 Subsumes: PRI-823 (Codex App compatibility)docs/superpowers/specs/2026-04-06-worktree-rototill-design.md
- Platform-neutral config-file references — Phase B designPhase A (see 2026-05-05-platform-neutral-prose-design.md) replaced generic third-person "Claude" prose with agent-neutral forms. This phase tackles the next category: references to the per-platform i…docs/superpowers/specs/2026-05-05-platform-neutral-config-refs-design.md
- Platform-neutral prose — Phase A designSuperpowers ships to multiple agent runtimes (Claude Code, Codex, Cursor, OpenCode, Copilot CLI, Gemini CLI). Skill content and supporting docs were written first for Claude Code and use "Claude" in…docs/superpowers/specs/2026-05-05-platform-neutral-prose-design.md
- Platform-neutral README ordering — Phase C designPhases A and B (see 2026-05-05-platform-neutral-prose-design.md and 2026-05-05-platform-neutral-config-refs-design.md) already neutralized generic Claude prose and config-file references in the READM…docs/superpowers/specs/2026-05-05-platform-neutral-readme-design.md
- Lift drill into superpowers as `evals/` — designDrill is a Python skill-compliance benchmark that lives in its own repo at obra/drill. It drives real tmux sessions, runs an LLM actor as a simulated user, runs an LLM verifier on the resulting trans…docs/superpowers/specs/2026-05-06-lift-drill-into-evals-design.md
- SDD Task-Scoped Review DispatchMake subagent-driven-development's per-task reviews cheaper and faster without weakening them, by scoping per-task review prompts to the task and stopping redundant work — while final branch review s…docs/superpowers/specs/2026-06-09-sdd-task-scoped-review-dispatch-design.md
- Positive-Instruction Redesign of Skill Guidance — Design SpecStatus: Proposed (follow-up to the 2026-06-09 SDD review-dispatch work; separate PR per the one-problem-per-PR rule) Driver: Measured evidence (2026-06-10) that some negative instructions in skill pr…docs/superpowers/specs/2026-06-10-positive-instruction-redesign-design.md
- Strict-Cost SDD — Design SpecStatus: Proposed experiment ladder (not implementation). Each rung ships only with its gate evidence; abort any rung whose gates fail. Objective: minimize dollars per plan-execution. Wall-clock is un…docs/superpowers/specs/2026-06-10-strict-cost-sdd-design.md
- Visual Companion Auth Hardening DesignDate: 2026-06-10 Status: Draft for Drew reviewdocs/superpowers/specs/2026-06-10-visual-companion-auth-hardening-design.md
- Visual Companion Final Hardening Fixup DesignDate: 2026-06-11 Status: Draft for Drew reviewdocs/superpowers/specs/2026-06-11-visual-companion-final-hardening-fixup-design.md
- SDD plan-scoped workspace — eval resultswith maintainer sign-off after the RED baseline did not reproduce blind stale-ledger adoption. 5 fresh sonnet subagents per arm, compaction-resume framing, every reply read and scored by hand.docs/superpowers/specs/2026-07-06-sdd-plan-scoped-workspace-eval-results.md
- SDD plan-scoped workspace — designSDD's durable-progress workspace (.superpowers/sdd/, introduced v6.0.0/v6.0.3) has no plan identity and no end-of-life. Every artifact is keyed by bare task number (progress.md, task-N-brief.md, task…docs/superpowers/specs/2026-07-06-sdd-plan-scoped-workspace.md
- SDD Fix-Loop Redesign — Design SpecStatus: Approved design (brainstormed with Jesse 2026-07-15); implementation plan to follow. Objective: make the subagent-driven-development skill's review-fix loop convergent and autonomous, and mak…docs/superpowers/specs/2026-07-15-sdd-fix-loop-redesign-design.md
- Codex Efficiency Fixes — DesignDate: 2026-07-30 Status: approved by Jesse (in-session) Branch: codex-efficiency-fixes off devdocs/superpowers/specs/2026-07-30-codex-efficiency-fixes-design.md
- Hermes Version-Bump Wiring DesignDate: 2026-08-05 Revised: 2026-08-06 Status: Approveddocs/superpowers/specs/2026-08-05-hermes-version-bump-wiring-design.md
- Diagnosing Superpowers Sessions — DesignDate: 2026-08-27 Status: approved by Jesse (in-session); spec pending review Branch: diagnosing-superpowers off devdocs/superpowers/specs/2026-08-27-diagnosing-superpowers-design.md
- Brainstorming rebuild: intentCoding agents are very good at building. They build what they believe the human wants, and they rarely ask enough or look around enough before they start. The brainstorming skill exists to close that…docs/superpowers/specs/2026-10-05-brainstorming-rebuild-design.md
- Antigravity Marketplace pluginList Superpowers in the Google Antigravity plugin Marketplace, as part of a partnership with Google. That requires a native Antigravity plugin manifest with the Marketplace presentation fields, a log…docs/superpowers/specs/2026-10-06-antigravity-marketplace-design.md