# GanyuanRan/Aegis

Make AI coding agents architecture-aware: baseline-first, evidence-verified, drift-checked, and safe across long tasks.

Repository: https://github.com/GanyuanRan/Aegis
Canonical: https://ross.abutalabs.com/products/ganyuanran-aegis
Homepage: https://github.com/GanyuanRan/Aegis
Language: Python
License: MIT
License Family: permissive
Topics: agent-skills, ai-agents, ai-coding, architecture-driven-development, claude-code, codex, coding-agents, evidence-driven, opencode, software-architecture, baseline-first, first-principles, awsome-coding-plugin, dive, spec-driven-development, spec-kit, dsh-plugin
Last push: 2026-09-02T15:15:53+00:00

## Health v2 (maintenance only)
Score: 77/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 100, release rhythm 88, longevity 8
- inputs: {"age_days": 125, "days_push": 0, "days_rel": 1, "gap_med": 0, "n_releases_24m": 108}
- flags: young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1165, forks 51 (observed 2026-09-03T02:15:05.968900+00:00)

## What it is
Aegis is a method pack (plugin/skills collection) that makes AI coding agents architecture-aware, baseline-first, evidence-verified, and drift-checked across long tasks. It ships as plugins for tools like Claude Code, Codex, Cursor, OpenCode, and Windsurf to make agent edits safer and more reliable.

## Use cases
- make my AI coding agent follow the codebase architecture
- stop my coding agent from breaking things in long tasks
- verify agent changes against a baseline before accepting them
- add spec-driven development skills to Claude Code or Codex
- reduce rework from AI-generated code
- detect drift when an AI agent edits my project

## When to choose
- you use AI coding agents like Claude Code, Codex, Cursor, or OpenCode on real codebases
- you want evidence-based 'done' criteria and drift checks for agent edits
- you want a method pack of agent skills rather than building your own prompts

## When to avoid
- you don't use AI coding agents in your workflow
- you need a standalone application rather than an agent plugin/method pack
- you want fully autonomous agents without baseline/evidence constraints

## Facets
- artifact type: plugin
- maturity: active
- function: agent-framework, prompt-engineering, developer-tools, workflow-automation
- domain: developer-tools
- platform: cli, cross-platform, python
- tags: ai-coding-agents, claude-code, codex, method-pack, spec-driven-development, architecture-driven-development, agent-skills, drift-checking, ai-agents, software-architecture, automation

## Member repositories
- GanyuanRan/Aegis (main) score 77

## Provenance
- Observed fields: from GitHub, fetched 2026-09-03T02:15:05.968900+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:37:23.930300+00:00, confidence not recorded.
  - readme: https://github.com/GanyuanRan/Aegis (fetched 2026-09-03T02:15:05.968900+00:00, sha 627632ca8079)
  - homepage: https://github.com/GanyuanRan/Aegis (fetched 2026-08-29T12:42:10.669457+00:00, sha 0c38e0eff74f)
- Data as of 2026-08-30T08:39:29.467469+00:00.