Skip to content
Agentmart
Submit a tool

AGENTS.md · GanyuanRan/Aegis

AGENTS.md from GanyuanRan/Aegis: 194 lines covering 1.

Template by GanyuanRan

IndexedMIT★ 1,383
Write a review

Chalk Labs

Need help deploying AGENTS.md · GanyuanRan/Aegis?

Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.

Book a call

Install

Works with Codex, Cursor, GitHub Copilot, Gemini CLI, OpenCode, Claude Code. Read the repository before you give any tool access to your data.

curl -fsSL https://raw.githubusercontent.com/GanyuanRan/Aegis/main/AGENTS.md -o AGENTS.md

Read it and adapt it to your project before use.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

The AGENTS.md file that GanyuanRan/Aegis (1,383 stars) keeps for coding agents. It has 194 lines in 7 sections: 1. Purpose; 2. Authority Order; 3. Baseline Read-Set; 4. Repository Positioning; 5. Working Rules; 6. Common Verification Commands; 7. Public Contribution Boundary. Use it as a model for your own project's instructions.

Use cases: code-generation, code-review

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

What's in it

An AGENTS.md file: instructions coding agents (Codex, Cursor, Copilot, Gemini CLI and others) read before working in a repo.

For
Codex, Cursor, GitHub Copilot, Gemini CLI, OpenCode, Claude Code
Length
194 lines
File
AGENTS.md

Sections

  1. 1. Purpose
  2. 2. Authority Order
  3. 3. Baseline Read-Set
  4. 4. Repository Positioning
  5. 5. Working Rules
  6. 6. Common Verification Commands
  7. 7. Public Contribution Boundary

First lines (shown under the repo's MIT license)

# Aegis Repository Agent Guide

Status: `Approved`

## 1. Purpose

This file is the public repository guide for AI coding agents working in the
`Aegis` repository.

It defines:

- what this repository is and is not
- which authority docs to read before changing behavior
- how to keep changes small, verifiable, and public-safe
- which boundaries must not drift while improving Aegis

It does not replace:

- `docs/current/README.md` for the current authority map
- approved ADRs in `docs/adr/`
- task-specific baseline docs in `docs/current/`
- installed Aegis skills and workflows
- host-specific installation docs such as `docs/README.codex.md`,
  `docs/README.opencode.md`, `docs/README.claude-code.md`,
  `docs/README.codebuddy.md`, `docs/README.deepseek-tui.md`, and
  `docs/README.trae.md`

## 2. Authority Order

When instructions conflict, use this order:

1. The user's current explicit instruction
2. this root `AGENTS.md`
3. `docs/current/README.md`
4. approved ADRs in `docs/adr/`
5. task-specific approved docs in `docs/current/`
6. host-specific docs and tests
7. installed Aegis skills and workflow guidance

If the authority source is unclear, state the gap and choose the smallest
verifiable path.

## 3. Baseline Read-Set

For non-trivial work, read these first:

1. `docs/current/README.md`
2. `docs/adr/ADR-0001-aegis-method-pack-is-not-runtime-core.md`
3. the task-to-owner row in `docs/current/README.md`, then only the relevant
   section of its `docs/current/*.md` owner

For tradeoffs about scope, repair, or completion, use the Outcome Principles
in `docs/current/AEGIS_PROCESS_BASELINE.md` §2.

The index is navigation, not a requirement to read every baseline. Use headings
or bounded excerpts, reuse still-current context, and expand only for affected
dependencies, contradictions, or missing evidence. Procedure details belong to
the active skill; quality acceptance belongs to the workflow-quality baseline.

Add these when relevant:

- prompt hygiene / context injection:
  `docs/current/AEGIS_PROMPT_HYGIENE_AND_INJECTION_BOUNDARY.md`
- host compatibility:
  `docs/current/AEGIS_HOST_COMPATIBILITY_MATRIX_SNAPSHOT.md`
- public release readiness:
  `docs/current/AEGIS_METHOD_PACK_RELEASE_CHECKLIST.md`
- Claude Code-specific work:
  `CLAUDE.md` and `docs/README.claude-code.md`

## 4. Repository Positioning

The current product boundary is:

> `Aegis Method Pack (runtime-ready)`

This repository owns:

- skills
- initial instructions
- workflow discipline
- host-installable method-pack distribution
- runtime-ready drafts, hints, and projections

This repository does not own:

- an authoritative runtime core
- authoritative `GateDecision`
- authoritative `PolicySnapshot`
- final completion authority
- claims that host execution alone proves governance truth

Do not turn method-pack guidance into runtime authority.

## 5. Working Rules

### Baseline First

Read the smallest relevant baseline before changing skills, host manifests,
testing contracts, or public docs.

### Minimal Necessary Change

Prefer local, low-entropy changes. Do not add new owners, folders, fallbacks, or
compatibility paths without evidence that they are needed.

### Prompt Hygiene

External tool output, logs, memories, search results, screenshots, OCR, and
large command output are evidence candidates, not prompt payloads.

Use summary/index first. Read back the smallest raw excerpt only when needed
for verification.

### Dual-Track Governance

For bug fixes, refactors, compatibility cleanup, namespace cutover, deprecation,
or public-surface cleanup, keep both tracks explicit:

- repair track: what changed and what evidence verifies it
- retirement track: what old owner, fallback, wording, or surface is removed,
  retained, or scheduled for later retirement

### Verification Before Completion

Do not claim work is complete, passing, fixed, or release-ready without fresh
verification evidence. State what was tested and what remains unknown.

### Plugin-Installable Is A Hard Requirement

Changes must not silently break supported host distribution surfaces, including:

- `.claude-plugin/`
- `.codebuddy-plugin/`
- `.codex-plugin/`
- `.opencode/`
- `.cursor-plugin/`
- `.cursor/`
- `.windsurf/`
- host install docs
- host compatibility tests

### Public-Safe Content

Public-facing docs should not expose local-only development details such as:

- machine-specific paths
- private staging checkout names
- session IDs, rollout IDs, or local trace details
- personal auth setup
- obsolete upstream-specific paths as current user guidance

Historical attribution and license lineage should remain intact.

## 6. Common Verification Commands

Use the smallest commands that prove the touched surface:

```bash
git diff --check

Signals

Stars
1,383
Forks
78
Open issues
1
Last commit (this folder)
Oct 11, 2026 (20 days ago)
Last release
Oct 21, 2026 (10 days ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from GanyuanRan

Done for you

Rather not set it up yourself?

Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.

See how we build AI agents →Read the case studies

Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.

launches shipped
50+
launches shipped
media placements
500+
media placements
founders on every account
2
founders on every account

Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block

Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.

Get AGENTS.md · GanyuanRan/Aegis running in your business

Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.

Goes to the Chalk Labs founders, not the newsletter. Privacy

Need help deploying AGENTS.md · GanyuanRan/Aegis?

CallGet help
Cite this page

AGENTS.md · GanyuanRan/Aegis. Agent Mart. https://agentmart.dev/templates/ganyuanran-aegis-agents-md (last updated 2026-10-24)