Skip to content

mcouthon/agents

Repository files navigation

AGENTS

AI-Guided Engineering — Navigate → Think → Ship

License: MIT PRs Welcome GitHub Discussions

A minimal framework for AI-assisted coding with phase-based workflows, auto-activating skills, and enforced tool safety. Works with VS Code Copilot and Claude Code.


What You Get

Component Count What It Does
Agents 5 Phase-based workflow with enforced tool restrictions and handoff buttons
Skills 6 Auto-activate based on your prompts (debug, mentor, architecture, etc.)
Instructions 5 File-type coding standards that load automatically
git clone https://github.com/mcouthon/agents.git
cd agents
./install.sh

That's it. Use /agent or the Chat menu to select agents, or just talk naturally and let skills auto-activate.


The Core Insight

"The highest leverage point is at the end of research and the beginning of the plan. A human can skim 30 seconds and provide feedback that saves hours of incorrect implementation."

This framework is built around that insight. The Explore agent is read-only—it can't accidentally edit your code. You review its research and plan, then hand off to Implement when you're ready.


The Workflow

Explore → Implement → Review → Commit
    ↓         ↑         ↓
  Handoff   (fix)     Done
Agent Purpose Tool Access Handoff To
Explore Research + create plans Read-only → Implement, → Handoff
Implement Execute planned changes Full access → Review
Review Verify implementation quality Read + Test → Commit / → Explore (re-plan)
Commit Create semantic commits Git + Read ✅ Done
Handoff Persist context for next session Write → Implement

Why agents? Each phase has enforced tool restrictions (Explore can't accidentally edit code) and handoff buttons to guide you to the next step.


Skills (Auto-Activate)

Skills activate automatically based on what you say:

You Say Skill Activated
"This test is failing" debug
"Find code smells" tech-debt
"Document the architecture" architecture
"Teach me how this works" mentor
"Clean up dead code" janitor
"Challenge my approach" critic

No manual switching required—just ask naturally.


What AGENTS Is / Isn't

AGENTS Is AGENTS Isn't
Advisory guidance Mandatory enforcement
Phase-based workflow Magic one-shot agent
Minimal and composable Batteries-included framework
IDE-agnostic patterns Cursor/Claude-specific
Human-in-the-loop at key points Fully autonomous

Installation Details

After ./install.sh:

Component Installed To
Agents (VS Code) ~/Library/Application Support/Code/User/prompts/
Commands (Claude Code) ~/.claude/commands/
Skills ~/.github/skills/ (with ~/.claude/skills symlink)
Instructions ~/Library/Application Support/Code/User/prompts/
Handoffs gitignore Added to global gitignore

Claude Code Usage

Agents are available as slash commands in Claude Code:

Command Purpose
/project:explore <task> Research and plan
/project:implement Execute the plan
/project:review Verify changes
/project:commit Create semantic commits

Example workflow:

$ claude
> /project:explore add user authentication

[Claude researches, produces plan]

> /project:implement

[Claude implements based on conversation context]

> /project:review

[Claude reviews changes]

> /project:commit

[Claude creates commits]

Note: VS Code agent features like tool restrictions, model selection, and handoff buttons are not available in Claude Code. Skills work identically on both platforms.


Customization

Adding an Agent

Create .github/agents/my-agent.agent.md:

---
name: My Agent
description: What this agent does and when to use it.
tools: ["codebase", "search", "editFiles"]
model: Claude Sonnet 4 # Optional
handoffs:
  - label: Next Step
    agent: other-agent
    prompt: Continue with the next phase.
---
# My Agent Instructions

Your detailed instructions here.

Adding a Skill

Create .github/skills/my-skill/SKILL.md:

---
name: my-skill
description: >
  Trigger keywords for auto-activation: "keyword1", "keyword2".
  Focus on WHEN to use (symptoms), not WHAT it does.
---
# My Skill Instructions

Your instructions here (< 500 lines recommended).

Validating Skills (TDD for Documentation)

  1. RED - Run task WITHOUT the skill, note failures
  2. GREEN - Add skill, verify improvement
  3. REFACTOR - If agent rationalizes around it, strengthen guidance

If you didn't see it fail without the skill, you don't know if the skill helps.

Run ./install.sh after adding agents or skills.


Agents vs Skills

Use Case Use
Need enforced tool restrictions Agent
Need handoffs between phases Agent
Want auto-activation from prompts Skill
Role-based workflow phases Agent
Specialized methodologies Skill

File Structure

.github/
├── agents/           # Workflow phases (Explore, Implement, Review, Commit, Handoff)
└── skills/           # Auto-activating capabilities (debug, mentor, etc.)

instructions/         # File-type coding standards
├── global.instructions.md
├── python.instructions.md
├── typescript.instructions.md
├── testing.instructions.md
└── terminal.instructions.md

docs/
├── synthesis/        # Core principles and framework analysis
└── research/         # Research Decision Records (RDRs)

Troubleshooting

Skills not auto-activating?

  1. Run ./install.sh to ensure symlinks exist
  2. Check ~/.github/skills/ for your skills
  3. Be more explicit: "Use debug mode to investigate..."

Need to uninstall?

./install.sh uninstall

Contributing

We welcome contributions! See CONTRIBUTING.md for guidelines.

Quick ways to contribute:


Further Reading

Topic Document
Core principles prevailing-wisdom.md
Framework analysis framework-comparison.md
Memory & session continuity memory-and-continuity.md
Research decisions docs/research/
12-Factor Agents docs/sources/12-factor-agents/

Why This Exists

Synthesized from multiple frameworks into something minimal and useful:

Model recommendation: Claude Opus 4.5 for heavy lifting. When Sonnet struggles, Opus delivers.

About

Collection of agentic coding instructions

Resources

License

Contributing

Stars

Watchers

Forks

Packages

No packages published

Languages