What is Forge AGENTS.md?

A set of repository instructions for AI coding agents working on Forge, an AI-first development system with an orchestrator and specialist roles. It explains which instruction files and shared project documents agents must read and follow.

Use it to configure AI agents working in the Forge repository, including task decomposition, specialist hand-offs, and shared project documentation.

It gives different coding tools a common operating guide for routing work, sharing project context, and respecting repository rules. This reduces inconsistent behaviour between agents.

How lionshilov/Forge configures Claude Code and Codex

Reuse

Borrowing it

Nothing to install: this file belongs to lionshilov/Forge. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file

curl -O https://raw.githubusercontent.com/lionshilov/Forge/main/AGENTS.md

Clone the repo

git clone --depth 1 https://github.com/lionshilov/Forge

Made for: Claude Code, Codex, OpenCode.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/instructions/lionshilov/forge/agents-md/github.svg)](https://agentmods.dev/instructions/lionshilov/forge/agents-md)

Your own site

<a href="https://agentmods.dev/instructions/lionshilov/forge/agents-md"><img src="https://agentmods.dev/badge/instructions/lionshilov/forge/agents-md/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/instructions/lionshilov/forge/agents-md"><img src="https://agentmods.dev/badge/instructions/lionshilov/forge/agents-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 584 This file is loaded in full into every session.

When invoked 584 The same file — it is already loaded in full.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5.1	$0.00584	$0.00584
Opus 5.5	$0.00234	$0.00234
Sonnet 5.5	$0.00117	$0.00117
Haiku 4.5	$0.00058	$0.00058

Measured today against content hash 0aa4ccf9a287, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

Forge AGENTS.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured today.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

AGENTS.md · 19 lines

How it starts

The opening of the file, as written. The whole thing — 19 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md

Machine-readable entry point for AI coding agents, following the agents.md convention. Claude Code reads CLAUDE.md instead — that file is authoritative. This file exists so other tools (Cursor, Codex, Windsurf, Zed, …) can run Forge too.

What this repository is

Forge is an AI-first development system: an Orchestrator prompt plus twelve specialist agent prompts (Product, Designer, Architect, Analyst, Security, iOS/Swift, Frontend-Web, Backend, ML/CV, QA, DevOps, Docs) that share memory through project_context/*.md files.

Operating rules (for any AI agent)

Adopt the Orchestrator role defined in ./CLAUDE.md. Read it first — it is the authoritative system prompt: core loop, routing rules, decomposition format, anti-patterns.
The Orchestrator never writes code. Decompose the request, then act as the right specialist by reading agents/<name>/CLAUDE.md into context and continuing in that role (Forge calls this "role-switch" dispatch).
Read shared context before any work: at minimum project_context/PRODUCT.md, ARCHITECTURE.md, CONVENTIONS.md, INTERFACES.md. They are the contract between agents.
Honor the phase gates. Gated context files open with status: front matter (template → draft → ready, or n/a). Before acting as a specialist, run bash .claude/hooks/forge-gate.sh check <agent> (or read the front matter yourself) — a non-zero exit names the files and owners to route to first. When you finish a context file you own, set its status. Skipping a gate is the user's decision, never yours.
QA review is mandatory. Every implementation output is reviewed against project_context/CONVENTIONS.md using the checklist in agents/qa/CLAUDE.md — in a fresh/isolated context if your tool supports it. QA is read-only: it reports issues, the originating specialist fixes them.
Keep the shared memory honest. After every completed task update project_context/PROGRESS.md; log every failure in project_context/ERRORS_LOG.md with root cause and prevention.
Tools with native subagents (Claude Code): all twelve agents are registered in .claude/agents/*.md and can be dispatched in isolated contexts, and a PreToolUse hook enforces rule 4 on every dispatch. Tools without subagents: emulate with role-switch (rule 2) — the loop still works, just sequentially.

Read the full file on GitHub · 19 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

today First seen · 19 lines · 584 tokens per session scan A 0aa4ccf9a287

Subscribe to this mod's changes

Forge AGENTS.md is an instructions file published in the GitHub repository lionshilov/Forge (5 stars, last pushed 1mo ago), licensed MIT. It adds 584 tokens to every session, about $0.0023 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-11-07.

Building agents LLM Multi-Agent ai-agents anthropic claude

Forge AGENTS.md next to 6 instructions from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
Forge AGENTS.mdlionshilov/Forge	Instructions file	5	1mo ago	MIT	A	584	Claude Code, Codex, OpenCode
vscode buildNext.instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	9,719	GitHub Copilot
next.js AGENTS.mdvercel/next.js	Instructions file	143k	today	MIT	A	7,461	Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	5,516	GitHub Copilot
langchain AGENTS.mdlangchain-ai/langchain	Instructions file	149k	today	MIT	A	4,680	Codex, OpenCode
vscode copilot-instructions.mdmicrosoft/vscode	Instructions file	195k	today	MIT	A	3,093	GitHub Copilot
deepseek-harness AGENTS.mddeepseek-ai/deepseek-harness	Instructions file	275k	3d ago	MIT	A	4,313	Codex, OpenCode

Forge: Instructions file for Claude Code