ai-literacy-superpowers plugin
A plugin with 41 skills and agents: harness engineering, a spec-to-integration pipeline and three loops that enforce constraints
- 123 ★ on the repository
- Last commit
Extended access
Why this level
- Hooks run in every session, read and check project files and can write reports into the repository
- Commands create and change project setup files, CI templates and logs
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add plugins/ai-literacy-superpowersDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- Hooks run in every session, read and check project files and can write reports into the repository
- Commands create and change project setup files, CI templates and logs
This is third-party code. Review the repository files before installing.
What it does
ai-literacy-superpowers sets up a project for AI agents in Claude Code and GitHub Copilot CLI: a HARNESS.md file of constraints, commands, hooks, templates and an agent team. The orchestrator drives a task through spec-writer, tdd-agent, implementation, code-reviewer and integration-agent. Gates sit between the stages. The advocatus-diaboli agent objects to the spec and to the code, and a person answers every objection in writing. While even one is pending, the gate stays shut. The review loop is capped at three rounds, after which the task goes to a person. Hooks run in every Claude Code session: they check constraints when a file is written, run markdownlint, scan for secrets with gitleaks and nudge you toward /reflect, /harness-audit and /harness-health. The repository is a marketplace that also carries the model-cards and diagnostic-legibility plugins.
Who it is for. Teams running a project with several AI agents that want rules written down in HARNESS.md and checked in CI.
Good fit when
- A new project starts with an agent team and needs CLAUDE.md, HARNESS.md and CI templates from one /superpowers-init command
- Before implementation you want written objections to the spec while no tests or code exist yet
- You need a weekly report on the state of project rules: /harness-health and /harness-audit
- After commits you want to log lessons in REFLECTION_LOG.md and promote good ones to AGENTS.md
Not a fit when
- A five-minute task: the gated chain with objections adds more ceremony than value
- A stack with no implementer of its own: you create the language-specific implementers yourself
- You need one small skill: the set is large, 41 skills and 40 commands, and it installs as a whole
Example request
Run /superpowers-init in this project, show which files appear, then run /harness-status.Limitations
The plugin is large: 41 skills, 22 agents, 40 commands and 11 hooks, and the hooks are active in every Claude Code session, so review them before enabling. No language-specific implementers ship with it; you create one per stack. The constraint check on file writes and markdownlint are advisory, while strict checks run in CI on a pull request. GitHub Copilot CLI runs commands as prompts rather than slash commands, so some features may differ. GitHub metadata shows no license; the README states Apache 2.0.
How to disable. Remove the plugin through the /plugin command in Claude Code, then delete the files /superpowers-init created in the project: HARNESS.md, REFLECTION_LOG.md, MODEL_ROUTING.md and the CI templates.
FAQ
How is /superpowers-init different from /harness-init?
`/superpowers-init` creates everything: CLAUDE.md, HARNESS.md, AGENTS.md, MODEL_ROUTING.md, REFLECTION_LOG.md, the agent team, skills, hooks and CI templates. `/harness-init` creates only HARNESS.md with starter constraints and cleanup rules.
Does the plugin work in GitHub Copilot CLI?
Yes. Add the marketplace with `copilot plugin marketplace add Habitat-Thinking/ai-literacy-superpowers` and install with `copilot plugin install ai-literacy-superpowers@ai-literacy-superpowers`. Commands appear there as prompts.
What does the advocatus-diaboli agent do?
It is read-only and writes an objection record: first against the spec, then against the finished code. A person closes each objection with a written disposition, and while any is pending the gate blocks further work.
How do I update the plugin?
Run `claude plugin update ai-literacy-superpowers@ai-literacy-superpowers`, then `/harness-upgrade`: it compares your HARNESS.md with the plugin template and offers new content to adopt.
What else is in the marketplace?
Three plugins: `ai-literacy-superpowers`, `model-cards` for researching models and drafting model cards, and `diagnostic-legibility` with an agent that builds and cross-checks models of a codebase. Most of the README covers the first.
Discussion
No questions or reviews yet.
Related
PluginsThe /better-harness command reviews how an AI agent's work is set up in a project: goals, checks, delivery and memory, and returns a prioritized fix report
Prompt kit and plugin for Codex and Claude Code that turn repeated agent mistakes into repository rules.
Turns a GitHub repo into a package with its own AI agent: an npx command, an MCP server, memory and a signed release
One hundred ready agent teams for Claude Code, each with specialists, an orchestrator skill and domain skills, in English and Korean
A Codex plugin that builds a team of subagents and skills for a project: an orchestrator splits tasks and checks the result.
AI Builder Club Skills
AI Builder Club — Skills
A Claude Code plugin marketplace to make a repo agent-ready and run compounding agent loops
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Repository stars | Main difference |
|---|---|---|---|
| ai-literacy-superpowers (this card) | no | 123 | |
| Superpowers | no | 316,679 | installs the base Superpowers skill set; this adds HARNESS.md, spec objections and CI checks |
| Claude Code Harness | no | 3,207 | a different rule and hook set for Claude Code; this one centres on three enforcement loops and gated agents |
README in short
A marketplace of three plugins, with the flagship covered in depth: lists of skills, agents, commands, templates and hooks, three enforcement loops, the agent pipeline, dynamic workflows and compound learning. Separate sections cover observability and the theory behind the approach, from architecture and literate programming to cognitive science.