claude-leverage
starsforks
How to install claude-leverage?
filip-podstavec/claude-leverage
$git clone https://github.com/Filip-Podstavec/claude-leverageMake any repo AI-first - write sustainable code from the start, or refactor a legacy codebase to prepare it for agent-driven development.Building blocks for Claude Code: subagents, slash commands, hooks, and workflow patterns. Copy what you need. A working developer's stack for Claude Code.
README.md
| 1 | <div align="center"> |
| 2 | |
| 3 | <img src="assets/banner.svg" alt="claude-leverage" width="640"> |
| 4 | |
| 5 | **Make any repo AI-first - write sustainable code from the start, or refactor a legacy codebase to prepare it for agent-driven development.** |
| 6 | |
| 7 | [](https://docs.anthropic.com/en/docs/claude-code) |
| 8 | [](https://developers.openai.com/codex) |
| 9 | |
| 10 | <sub>Convention layer (<code>AGENTS.md</code>) is also read by any AGENTS.md-aware tool - Cursor, Windsurf, Copilot…</sub> |
| 11 | |
| 12 | <br> |
| 13 | |
| 14 | [](https://github.com/Filip-Podstavec/claude-leverage/actions/workflows/ci.yml) |
| 15 | [](LICENSE) |
| 16 |  |
| 17 |  |
| 18 |  |
| 19 |  |
| 20 | |
| 21 | </div> |
| 22 | |
| 23 | > A small, opinionated **AI-coding-agent suite** for Claude Code and Codex CLI: |
| 24 | > deterministic security hooks, cross-tool skills, and self-maintaining |
| 25 | > conventions that keep a repo safe for the *next* agent to modify. |
| 26 | > |
| 27 | > **Complements** skills-based plugins like the official `superpowers` plugin; |
| 28 | > it does not try to replace them. |
| 29 | |
| 30 | ## Contents |
| 31 | |
| 32 | - [The problem it solves](#the-problem-it-solves) |
| 33 | - [Who it's for](#who-its-for) |
| 34 | - [Quick start](#quick-start) |
| 35 | - [Install](#install) |
| 36 | - [What you get](#what-you-get) |
| 37 | - [How it works](#how-it-works) |
| 38 | - [Benchmarks](#benchmarks) |
| 39 | - [How it compares](#how-it-compares) |
| 40 | - [What's inside](#whats-inside) |
| 41 | - [Why one repo for two tools](#why-one-repo-for-two-tools) |
| 42 | - [Honest history](#honest-history) |
| 43 | - [Verify, update, uninstall](#verify-update-uninstall) |
| 44 | - [License](#license) |
| 45 | |
| 46 | ## The problem it solves |
| 47 | |
| 48 | Shipping at velocity with AI agents is easy. Shipping in a way that the *next* |
| 49 | agent (human or AI) opening the repo in six months can still safely modify is |
| 50 | the hard part. That second part is what the hooks, conventions, and skills here |
| 51 | automate - not just at session start, but continuously as the repo grows. |
| 52 | |
| 53 | <table> |
| 54 | <tr><th align="left">Without a stack</th><th align="left">With claude-leverage</th></tr> |
| 55 | <tr> |
| 56 | <td>Context dies between sessions - every agent re-learns the repo from scratch.</td> |
| 57 | <td>AIDEV anchors + the <code>context-surface</code> hook surface the right facts at the moment of edit.</td> |
| 58 | </tr> |
| 59 | <tr> |
| 60 | <td>Docs drift out of sync with the code until nobody trusts them.</td> |
| 61 | <td>Self-maintenance nudges flag missing <code>AGENTS.md</code>, stale anchors, and code↔docs drift while it's still cheap to fix.</td> |
| 62 | </tr> |
| 63 | <tr> |
| 64 | <td>Big tasks lose the plot; the <em>why</em> behind decisions evaporates.</td> |
| 65 | <td>Durable memory: ADRs and distilled session logs carry intent across sessions.</td> |
| 66 | </tr> |
| 67 | <tr> |
| 68 | <td>Nothing stands between a careless agent and a leaked secret or a force-push.</td> |
| 69 | <td>Deterministic shell hooks block secrets, force-push, and <code>--no-verify</code> before any model decision.</td> |
| 70 | </tr> |
| 71 | </table> |
| 72 | |
| 73 | Three properties guide every decision in this repo: |
| 74 | |
| 75 | 1. **Security by default** - deterministic hooks, no model needed. |
| 76 | 2. **Self-maintaining as the repo grows** - non-blocking nudges surface |
| 77 | maintenance debt early. |
| 78 | 3. **Cross-tool by design** - same `AGENTS.md`, same `SKILL.md`, same hook |
| 79 | scripts in Claude Code and Codex. Author once. |
| 80 | |
| 81 | ## Who it's for |
| 82 | |
| 83 | - **Freelancer on client repos** - you ship features across many client |
| 84 | codebases through agents, and each one has to stay safe to hand back. The |
| 85 | security hooks and `AGENTS.md` conventions travel with you. |
| 86 | - **Team inheriting a legacy repo** - you've opened a six-month-old codebase |
| 87 | with no docs. `/repo-doctor` scores its AI-readiness and names the |
| 88 | highest-leverage gaps; the bootstrap skills make it navigable fast. |
| 89 | - **Solo builder shipping at velocity** - you want speed without the repo |
| 90 | rotting into unmaintainable AI-slop. Self-maintenance nudges plus ADR/session |
| 91 | memory keep intent intact. |
| 92 | |
| 93 | ## Quick start |
| 94 | |
| 95 | **Inheriting or already have a repo? Start here.** Install the plugin: |
| 96 | |
| 97 | ``` |
| 98 | /plugin marketplace add Filip-Podstavec/claude-leverage |
| 99 | /plugin install claude-leverage@filip-podstavec |
| 100 | ``` |
| 101 | |
| 102 | Then make the repo AI-ready, top-down: |
| 103 | |
| 104 | 1. **`/repo-doctor`** - read-only AI-readiness audit. Scores ~20 dimensions |
| 105 | (Foundation / Why / What / In-code / Hygiene / Sync) and names the gaps that |
| 106 | matter most. |
| 107 | 2. **Fix the top gaps it reports** - usually `/init-repo` (drop in a root |
| 108 | `AGENTS.md`), then `/arch-map` and `/glossary-init` for machine-readable |
| 109 | module metadata and domain terms. |
| 110 | 3. **`/refresh-context-map`** - build the manifest that powers just-in-time |
| 111 | context surfacing, so the right anchors appear at edit time. |
| 112 | |
| 113 | **Starting a fresh repo?** Install, th |
People who liked this also viewed
6 picksType
- claude-codeClaude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflows - all through natural language commands.WorkflowsJul 2026
- TrellisThe best agent harness.WorkflowsJul 2026
- reflyThe first open-source agent skills builder. Define skills by vibe workflow, run on Claude Code, Cursor, Codex & more. Build Clawdbot 🦞· APIs for Lovable · Bots for Slack & Lark/Feishu · Skills are infrastructure, not prompts.WorkflowsMar 2026
- pyspurA visual playground for agentic workflows: Iterate over your agents 10x fasterWorkflowsJun 2026
- claude-code🚀 Open source Claude Code CLI source code. Advanced AI Agent for developers. Includes TypeScript codebase for LLM tool-calling, agentic workflows, and terminal UI. Remember this is just the skeleton not the brain itself. Found by Chaofan Shou. WorkflowsApr 2026
- claude-codeClaude Code is an agentic coding tool that lives in your terminal, understands your codebase, and helps you code faster by executing routine tasks, explaining complex code, and handling git workflows - all through natural language commands.WorkflowsApr 2026
