Agent Context Framework — AI skill for Claude Code
Production-ready Markdown (.md) context templates for AI coding agents (Cursor, Claude Code, Aider, Windsurf).
How to install Agent Context Framework
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open mhadikz/agent-context-framework and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Agent Context Framework does
Production-ready Markdown (.md) context templates for AI coding agents (Cursor, Claude Code, Aider, Windsurf). Eliminate AI hallucinations, prevent scope creep, and save context tokens.
Alternatives in AI
- Define — Definition phase - Clarify and scope problems with multi-AI consensus 2.8k ★
- Claudekit — by Carl Rannaberg - Impressive CLI toolkit providing auto-save checkpointing, code quality hooks, specificatio 633 ★
- The Pair — Open-source AI pair programming for desktop: a Mentor + Executor agent cross-check each other's code to catch 358 ★
README
🤖 AI Context Framework Templates
| 📝 License | 🤝 Contributions | 🤖 Target |
|---|---|---|
MIT |
PRs Welcome |
AI Coding Agents |
A collection of 10 industry-standard Markdown (`.md`) blueprints designed to act as an "onboarding manual" for AI coding agents and LLM assistants. By placing these files in your project root, you dramatically reduce AI hallucinations, prevent scope creep, and save thousands of context window tokens.
🚀 Why Use This Framework?
When working with AI tools like **Cursor, Claude Code, Aider, Windsurf, or GitHub Copilot**, LLMs often suffer from "context rotting" or forget project boundaries over long chat sessions.
This framework solves that by dividing project context into **three specialized pillars**:
- Instructions & Rules: Tell the AI how to behave and execute commands.
- Project Shaping: Keep the AI aligned with your roadmap and active sprints.
- Architecture & Governance: Stop the AI from rewriting your tech stack or breaking code style.
📂 What's Inside the Box?
🛠️ Pillar 1: Core AI Instructions & Rules
- `AGENTS.md` - The Universal AI Directive: The platform-agnostic standard defining project context, build commands, and safety guardrails.
- `CLAUDE.md` / `GEMINI.md` - Tool-Specific Overrides: Rules optimized for terminal-based CLI agents to dictate how diffs and outputs are formatted.
- `SKILL.md` - Deterministic Runbooks: Step-by-step workflow documentation that tells the AI exactly how to execute routines safely.
📋 Pillar 2: Project Shaping & Management
- `PLAN.md` - The Macro Blueprint: Sets the high-level vision, core milestones, and strictly defines what is out of scope.
- `BACKLOG.md` - The Task Queue: A clean Now/Next/Later pipeline that the AI can read to pull its next assignment autonomously.
Related Skills
Leo
A deterministic context-driven agentic SDLC framework for Cursor and Claude. 41 codified laws and 22 specialis
Agents Md Templates
One AGENTS.md, every AI coding agent. Copy-paste templates for Claude Code, Codex, Cursor, Gemini CLI, Copilot
Overscope
Catch AI coding-agent scope creep before you commit. Overscope compares Claude Code and Codex sessions with yo
Red Line Skill
A skill for AI agents (Claude, ChatGPT, etc.): clarify scope before planning, lay out an explicit end-to-end p
Bundlebox
The zero-token software factory for AI coding agents: scan, pack, route and run work across Claude Code, Codex
Vomit
Clean up Claude 5's token vomit with a separate LLM. Save your tokens, Opus is hopeless
Related Agents
10x Tool Calls
Cursor and Windsurf meter usage by requests and tool calls rather than tokens, which means a finished or stall
RAG Coder
Context-aware coder that grounds all code generation in actual codebase using RAG to eliminate hallucinations
Plan Challenger
Adversarial plan review before implementation. Attacks plans across 5 dimensions (Assumptions, Missing Cases,