cleburn

cleburn / aegis

Public

Open specification for AI agent governance. A structured operating contract for any AI agent working in your codebase — scoped roles, enforced permissions, autonomy levels, and coordination protocols. Schema-validated, machine-parseable, agent-agnostic. Apache 2.0.

26
8
100% credibility
Found Mar 11, 2026 at 23 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
AI Summary

Aegis is a conversational tool that scans a software project and generates structured, machine-readable policy files defining governance rules for AI agents.

How It Works

1
🔍 Discover Aegis

You hear about Aegis, a helpful guide that sets clear rules for AI assistants working on your project, making sure they follow your vision perfectly.

2
📥 Get Aegis ready

You bring Aegis onto your computer so it's all set to help with your project.

3
💬 Chat with Aegis

Aegis looks over your project and starts a friendly conversation, asking smart questions about your goals, boundaries, and how you want things done.

4
🗣️ Share your rules

You answer Aegis's questions about what's important, what AI helpers can touch, and how they should work together.

5
📋 Your rulebook appears

Aegis creates a clear set of files with all your rules, ready for any AI helper to read and follow without confusion.

6
🔍 Check and understand

You can ask Aegis to explain the rules in simple words or make sure everything matches what you said.

✅ AI helpers in sync

Now every AI assistant on your project knows the rules, works smoothly together, and builds exactly what you want.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 23 to 26 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is aegis?

Aegis is a TypeScript CLI that generates a .agentpolicy/ directory of structured JSON governance for AI agents in your codebase—like .editorconfig or dotnet editorconfig github, but for agent roles, permissions, autonomy levels, and task coordination. Run `aegis init` to scan your repo, chat conversationally via Anthropic Claude about principles and scopes, and get schema-validated files any agent can parse without retraining. It fixes agents rebuilding context from scratch or clashing in multi-agent flows.

Why is it gaining traction?

Unlike prose rules in READMEs or c# editorconfig github workarounds, Aegis outputs machine-queryable contracts with CLI tools like `aegis explain` for plain-language summaries and `aegis validate` for schema checks. The agent-agnostic format works with Claude Code, Cursor, or custom setups, and conversational setup feels natural, not bureaucratic.

Who should use this?

Devs running multi-agent workflows in TypeScript, .NET, or C# repos who need scoped roles for frontend/backend/testing without merge conflicts. Solo coders using Cursor for delegated code mods but wanting conservative guards on infra changes.

Verdict

Promising spec at 17 stars and 1.0% credibility score, with strong docs and examples offsetting early maturity—no tests visible yet. Try it globally via npm for AI-orchestrated projects; adopt the format even without the CLI.

(178 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.