agentleFS
Sign inStart free
GabrieleDattileGabrieleDattile/agent-superpowers/skills/orchestration/using-superpowers/SKILL.md

using-superpowers

Skill1 starsChanged 15 days agoMITView on GitHub

The master orchestration meta-skill for Agent Superpowers. Governs the 7-stage autonomous engineering lifecycle across Claude Code, Cursor, Codex, Antigravity, and Gemini CLI.

171 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use chinese-code-review, chinese-commit-conventions and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • Leodorareluctant259Skill

    chinese-code-review

    中文代码审查规范——在保持专业严谨的同时，用符合国内团队文化的方式给出有效反馈

    Also in 8 projects that use this

  • Leodorareluctant259Skill

    chinese-commit-conventions

    中文 Git 提交规范 — 适配国内团队的 commit message 规范和 changelog 自动化

    Also in 8 projects that use this

  • obraSkill

    diagnosing-superpowers

    Use when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.

    Also in 19 projects that use this

  • kntismSkill

    condition-based-waiting

    Use when tests have race conditions, timing dependencies, or inconsistent pass/fail behavior - replaces arbitrary timeouts with condition polling to wait for actual state changes, eliminating flaky tests from timing guesses

    Also in 3 projects that use this

  • frankxaiSkill

    defense-in-depth

    Use when invalid data causes failures deep in execution, requiring validation at multiple system layers - validates at every layer data passes through to make bugs structurally impossible

    Also in 3 projects that use this

  • MicrosoftSkill

    root-cause-tracing

    Use when errors occur deep in execution and you need to trace back to find the original trigger - systematically traces bugs backward through call stack, adding instrumentation when needed, to identify source of invalid data or incorrect behavior

    Also in 3 projects that use this

  • +3 more from Leodorareluctant259

The file

We split this skill into the rules it gives an agent

11 rules. 0 are common, 5 are distinctive, and 6 are very specific to this project.

  • 0

    Common

    Most of the 157 edited versions we read keep these.

  • 5

    Distinctive

    This author's own choices. Worth a look.

  • 6

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    SPECIAL TRIGGER: When the user asks:

    line 14
  2. Distinctive:

    Rule: Never jump into code if the prompt leaves architectural ambiguity.

    line 34
  3. Distinctive:

    Ask up to 3 clarifying questions, define acceptance criteria, and establish edge-case bounds.

    line 35
  4. Distinctive:

    Rule: Never experiment on the user's primary working tree for multi-file tasks.

    line 38
  5. Very specific:

    Create an ephemeral worktree branch via skills/workflow/git-worktree-sandbox/scripts/worktree.py create <task-name>.

    line 39
  6. Very specific:

    Run arch-archaeologist to map caller-callee graphs and dead functions.

    line 42
  7. Very specific:

    Run token-squeezer in interface mode to inspect large dependencies at 60% token reduction.

    line 43
  8. Distinctive:

    STRICT PROHIBITION: DO NOT TOUCH SOURCE CODE UNTIL A FAILING TEST IS WITNESSED.

    line 46
  9. Very specific:

    Write an isolated minimal test script (MRE).

    line 47
  10. Very specific:

    Run adversarial-duelist against your modified functions.

    line 53
  11. Very specific:

    Verify ghost-runner is GREEN.

    line 58

“Common” is measured: kept by at least 6 in 10 of the 157 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

⚡ Master Meta-Skill: Using Agent Superpowers

1. When to Trigger (Routing)

  • TRIGGER ALWAYS: This is the bootstrap skill. It should remain active across all agent sessions.
  • SPECIAL TRIGGER: When the user asks:
    • "What are your superpowers?"
    • "Tell me about your superpowers"
    • "Run diagnostics"
    • Respond by listing all active skills, their versions, and current operational readiness.

2. The 7-Stage Autonomous Engineering Lifecycle

Whenever tasked with building a feature, fixing a bug, or refactoring an architecture, you MUST orchestrate your skills through this disciplined lifecycle:

[1. BRAINSTORM] ──> [2. WORKTREE ISOLATE] ──> [3. MAP & COMPRESS]
       │                                              │
       ▼                                              ▼
[7. MERGE / CLEANUP] <── [6. RED-TEAM FUZZ] <── [5. SURGICAL PATCH] <── [4. TDD PROOF (RED)]

Stage 1: Brainstorming & Specification (feature-brainstormer)

  • Rule: Never jump into code if the prompt leaves architectural ambiguity.
  • Ask up to 3 clarifying questions, define acceptance criteria, and establish edge-case bounds.

Stage 2: Workspace Isolation (git-worktree-sandbox)

  • Rule: Never experiment on the user's primary working tree for multi-file tasks.
  • Create an ephemeral worktree branch via skills/workflow/git-worktree-sandbox/scripts/worktree.py create <task-name>.

Stage 3: Tomography & Context Compression (arch-archaeologist + token-squeezer)

  • Run arch-archaeologist to map caller-callee graphs and dead functions.
  • Run token-squeezer in interface mode to inspect large dependencies at 60% token reduction.

Stage 4: Test-Driven Proof of Defect (ghost-runner)

  • STRICT PROHIBITION: DO NOT TOUCH SOURCE CODE UNTIL A FAILING TEST IS WITNESSED.
  • Write an isolated minimal test script (MRE). Verify it fails with exit code != 0 (RED).

Stage 5: Surgical Implementation

  • Apply the minimal necessary edits strictly within the isolated worktree.

Stage 6: Adversarial Red-Team Fuzzing (adversarial-duelist)

  • Run adversarial-duelist against your modified functions.
  • Pass nulls, boundary overflows, SQL vectors, and zero divisions.
  • Code merges ONLY if it achieves a 100% defense score.

Stage 7: Verification & Worktree Cleanup

  • Verify ghost-runner is GREEN.
  • Merge the worktree back or prepare a clean Pull Request, then remove the ephemeral worktree.

3. Autonomous Arsenal Inventory

SkillRoleWhen Active
using-superpowers🧠 Master OrchestratorAlways active (boots the lifecycle)
feature-brainstormer💡 Requirements GuardWhen requirements are vague or a feature starts
git-worktree-sandbox📦 Environment IsolationMulti-file features or high-risk edits
arch-archaeologist🗺️ Code CartographerExploring unfamiliar/legacy codebases
token-squeezer📉 Context OptimizerBefore loading large files/dependencies into memory
ghost-runner🔬 TDD Defect ProverAny bug fix or regression remediation
adversarial-duelist⚔️ Offensive Red-TeamerPre-commit security and edge-case audit
api-synthesizer🔌 Network-to-SDK EngineIntegrating endpoints, cURL, or HAR traffic

4. Guardrails & Negative Constraints

  • ❌ NEVER bypass Stage 4 (TDD Proof) when fixing defects.
  • ❌ NEVER skip red-team fuzzing for public endpoints or auth logic.
  • ❌ NEVER leave orphaned git worktrees behind after task completion.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free