CLOUDM0N
Repository intelligence

obra/superpowers

GitHub

A software development methodology for coding agents built on top of a set of composable skills and instructions that ensure the agent executes them.

CLOUDM0N decision
REVIEW BEFORE ADOPTION
Trust REVIEW · 80/100
Good fit if

Developers using AI coding agents who want to enforce structured, test-driven development (TDD) and systematic debugging

Watch out for

Must be installed and configured separately for each coding agent harness in use

Practical intelligence

What matters before you adopt it

Problem it solves

AI coding agents prematurely jumping into writing code without structured planning, resulting in poor design, lack of testing, and unstructured debugging.

Best for
Developers using AI coding agents who want to enforce structured, test-driven development (TDD) and systematic debugging
Teams using supported agent harnesses (such as Claude Code, Cursor, Devin, or Gemini CLI) seeking consistent, multi-hour autonomous development behaviors
Main trade-offs
Must be installed and configured separately for each coding agent harness in use
Hermes Agent sessions lose their bootstrap if compaction occurs on the first turn, requiring a session restart
Contributions of new skills from the community are generally not accepted
Why it stands out
Supports over a dozen major AI agent harnesses and command-line interfaces
Strictly enforces mandatory workflows (like writing tests first and deleting code written before tests) rather than treating them as optional suggestions
Provides a robust library of pre-built skills for git workspace isolation, TDD, planning, and systematic 4-phase debugging
Trust & CVEs

Security evidence without the noise

Trust remains a decision signal; CVEs and scanner evidence explain what is driving the risk.

Security findings
2
CLOUDM0N scanner findings
Critical
0
High
0
Medium
2
Low
0
View trust evidence & security findings▼
Why this score
No trust rationale was stored for this scan.
CLOUDM0N findings
MEDIUM
Dynamic code execution pattern detected. 3 sample match(es) found.
MEDIUM
The project can spawn operating-system processes; review command construction and input handling. 5 sample match(es) found.
Architecture from code7 modules · 1 edges▼
Structural evidence

Modules and dependency edges extracted from repository code. This is code evidence, not README inference.

Code files
18
Modules
7
Dependency edges
1
Core modules
skills
4 files
Dependency flow
No dependency edges were extracted.
Detected languages
Shell · JavaScript · Python · TypeScript
Detected frameworks
—
Architecture evidence details
flowchart TD
    %% superpowers — high-level architecture (DRAFT, refine me)
    n0["skills · 4 files"]
    n1["tests · tests · 14 files"]
    n1 --> n0
    class n1 test
    classDef test fill:#499894,color:#ffffff,stroke:#397975
Evidence, security & integrations▼
Integrations
Claude CodeAntigravityCodex AppCodex CLICursorDevin CLIFactory DroidGemini CLI
Security notes
• Visual companion telemetry does not include details about projects, prompts, or coding agents, and can be completely disabled by setting SUPERPOWERS_DISABLE_TELEMETRY to true
Still unknown
• The README does not provide technical specifications for local execution, such as minimum hardware or software runtime versions
• Does not clarify the exact network requests or logging details collected on the visual companion server
• Does not detail how parallel agent dispatching is configured or whether it incurs separate API token costs
Adoption guidance▼
Adopt if
+ You want your AI coding agent to follow strict test-driven development methodologies autonomously
+ You are using supported agent harnesses like Claude Code or Cursor and want them to stick to structured plans without deviating over multi-hour runs
Avoid if
– You prefer coding agents that immediately generate code on demand without interactive Socratic specification phases
– You are using unsupported agent frameworks or require a platform where you can easily contribute custom development skills back to the main repository
How it works & getting started▼
How it works
1.Agent intercepts startup to brainstorm and refine specifications with the user instead of immediately writing code
2.Agent breaks down the approved design into a clear, bite-sized implementation plan (2-5 minutes per task)
3.Agent dispatches subagents to work through tasks autonomously using a two-stage review process (spec compliance, then code quality)
4.Agent enforces test-driven development (RED-GREEN-REFACTOR) and deletes code written prior to tests
5.Agent executes a code review against the plan before finishing and cleaning up the workspace
Getting started
→ For Claude Code: Install the plugin from Anthropic's official marketplace or register and install via the Superpowers marketplace
→ For Codex App: Open Plugins in the sidebar, find Superpowers in the Coding section, and click the '+' icon
→ For Cursor: Search and install from the plugin marketplace inside Cursor Agent chat
→ For Hermes Agent: Install as a Hermes plugin from the repository and restart any active sessions
Agent handoff
Use with any agent
JSON API
Alternatives

Nearby repositories worth comparing before adoption.

Compare top options →
CowAgent
zhayujie/CowAgent
64
Fit

Open-source super AI assistant & Agent Harness. Plans tasks, runs tools and skills, self-evolves with memory and knowledge. Multi-model, multi-channel. Lightweight, extensible, one-line install. (formerly chatgpt-on-wechat)

Trust REVIEW · 35
Compare →
cc-switch
farion1231/cc-switch
63
Fit

A cross-platform desktop All-in-One assistant for Claude Code, Codex, OpenCode, OpenClaw, Grok Build & Hermes Agent. Only official website: ccswitch.io

Trust REVIEW · 55
Compare →
dify
langgenius/dify
62
Fit

Build Agentic workflows, RAG pipelines, with rich AI model and tool support on one collaborative workspace. Deploy on cloud, VPC, or self-hosted, so teams move from prototype to production without rebuilding the stack.

Trust REVIEW · 10
Compare →