A comprehensive, vendor-agnostic framework for consistent AI-assisted development workflows - standardized instructions and commands that work seamlessly across Claude, Gemini, Copilot, and local AI models. See also: https://github.com/JacobPEvans/claude-code-plugins
SaferSkills independently audited ai-assistant-instructions (Agent Skill) and scored it 96/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 1 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 1 flagged
The text {match} tells the agent to skip the normal "ask the user first" gate. Used adversarially it removes the human-in-the-loop check before destructive or sensitive actions, turning a normally-gated agent into a fire-and-forget executor.
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Teaching AI assistants how to help you better. Yes, it's AI instructions written with AI assistance. We've come full circle.
[![License][license-badge]][license-url] [![Markdown Lint][markdownlint-badge]][markdownlint-url] [![pre-commit][precommit-badge]][precommit-url]
A centralized collection of instructions, workflows, and configurations for AI coding assistants. Drop these into your projects and get consistent, high-quality AI assistance across Claude, Copilot, and Gemini.
Think of it as a style guide, but for your AI pair programmer.
This repository maintains the generic, plugin-independent pieces:
AGENTS.md / CLAUDE.md / GEMINI.md configurationagentsmd/rules/agentsmd/workflows/Boundaries between this and the other AI-configuration layers (what lives here versus what is plugin-delivered or published to the docs site) are documented at docs.jacobpevans.com/ai-development/repo-boundaries.
gh) 2.0+ (for PR/issue management)# 1. Clone the repo
git clone https://github.com/JacobPEvans/ai-assistant-instructions.git
# 2. Copy AGENTS.md into your project
cp ai-assistant-instructions/AGENTS.md your-project/
# Optional: copy the auto-loaded rules too
mkdir -p your-project/agentsmd
cp -r ai-assistant-instructions/agentsmd/rules your-project/agentsmd/
# 3. Create vendor symlinks so each AI tool reads the same source
cd your-project
ln -sf AGENTS.md CLAUDE.md
ln -sf AGENTS.md GEMINI.mdOr just browse the documentation and cherry-pick what you need.
This repository ships configuration only — no slash commands, skills, agents, or hooks. Those are delivered as a separate, installable Claude Code plugin marketplace. Once the marketplace is added and the plugins are enabled, commands referenced by the workflow (for example /refresh-repo, /finalize-pr, /ship) become available alongside the configuration in this repo.
Once installed, the AI assistants read CLAUDE.md / AGENTS.md / GEMINI.md automatically at session start, and the auto-loaded rules in agentsmd/rules/ are pulled in for every session.
See the default workflow below for the expected development loop, and AGENTS.md for the full set of rules, routing decisions, and on-demand standards.
.
├── AGENTS.md # Canonical configuration (GEMINI.md is a symlink)
├── CLAUDE.md # Stub — Nix wiring auto-loads AGENTS.md globally; no re-import
├── agentsmd/
│ ├── rules/ # Auto-loaded universal and path-scoped rules
│ ├── workflows/ # Default workflow plus full-discipline guidance
│ └── docs/ # Workflow and integration support docs
├── .copilot/instructions.md # Symlink → AGENTS.md
├── .gemini/config.yaml # Gemini-specific config
├── scripts/ # Validation helpers (token limits, links)
└── .github/workflows/ # CI gates (markdown, spellcheck, link check, CodeQL, release-please)| Assistant | Integration | Notes |
|---|---|---|
| Claude | .claude/ directory | Full command support via Claude Code |
| GitHub Copilot | .github/copilot-instructions.md + prompts | Works in VS Code, GitHub.com, Visual Studio |
| Gemini | .gemini/ directory | Style guide and config support |
This repo centers on a lightweight autonomous loop:
Full PRD/test-first discipline remains available on demand for high-risk, multi-session, compliance-sensitive, cross-owner, or explicitly gated work. Details live in agentsmd/workflows/.
The documentation covers:
Browse agentsmd/rules/ and agentsmd/docs/.
Contributions welcome. See CONTRIBUTING.md for the details, though the short version is: open a PR, don't be a jerk, and I'll probably merge it.
Found a vulnerability? Please report it responsibly. See SECURITY.md for details.
Apache 2.0 - Use it, modify it, just keep the attribution.
[license-badge]: https://img.shields.io/badge/License-Apache_2.0-blue.svg [license-url]: LICENSE [markdownlint-badge]: https://github.com/JacobPEvans/ai-assistant-instructions/actions/workflows/markdownlint.yml/badge.svg [markdownlint-url]: https://github.com/JacobPEvans/ai-assistant-instructions/actions/workflows/markdownlint.yml [precommit-badge]: https://img.shields.io/badge/pre--commit-enabled-brightgreen?logo=pre-commit [precommit-url]: https://github.com/pre-commit/pre-commit
Part of a larger ecosystem of ~40 repos — see how it all fits together.
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.