Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add instructions/srmackey/context-forge/agents-mdgit clone --depth 1 https://github.com/srmackey/context-forgeWhat 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 | $0.02834 | $0.02834 |
| Opus 5 | $0.01417 | $0.01417 |
| Sonnet 5 | $0.00567 | $0.00567 |
| Haiku 4.5 | $0.00283 | $0.00283 |
Grade A, and why
context-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 yesterday.
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.
How it starts
The opening of the file, as written. The whole thing — 162 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
This file provides guidance to AI coding agents (Cursor, Claude Code, Grok Build, etc.) when working on the ContextForge codebase.
Agent Perspective
When working on or evolving ContextForge, bring an expert, collaborative perspective:
You serve as a careful steward and expert advisor for a lightweight, durable, git-native context substrate.
ContextForge's purpose is to provide reusable, cross-repo engineering knowledge in a form that remains valuable primarily through ordinary markdown files — even without AI assistance. The role is to offer high-quality guidance that strengthens this foundation thoughtfully, introducing change primarily when real usage demonstrates the need.
Principles for Careful Evolution
- Markdown remains the foundation. The raw
.mdfiles and their structure should stay the most important interface. Evolution should preserve or improve their readability, editability by humans, grep-ability, and git-friendliness. - Favor explicit mechanisms. Clear, single-purpose operations and explicit connections (such as
resolve_ref,link_task,import_contentpaired withrefresh_source, and governance references) tend to produce more reliable and understandable behavior than hidden automation. It is often wise to make important relationships visible and intentional. - Keep the system intentionally small until usage justifies growth. The current shape — filesystem as source of truth, SQLite as a derived index, FTS5, stdio transport, four entity types — reflects deliberate choices to avoid premature complexity. New capabilities or abstractions should be introduced primarily when actual usage (instrumented via logs,
droppedlists, trial feedback, or repeated friction) shows they are needed. Most promising ideas are best left for later. - Design for durability and graceful degradation. The system should continue to deliver value if the index needs rebuilding or if the MCP server is not running. Expert judgment usually avoids creating strong dependencies on the live server or AI presence.
- Protect development boundaries. Artifacts related to the system's own development and history belong in
.system/. Keeping them isolated helps maintain clarity about what is part of the product versus what is process. - Consider dogfooding as one useful signal. Using ContextForge during work on the project itself can surface practical insights.
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.
- yesterday First seen · 162 lines · 2,834 tokens per session scan A d7a67e65c956
context-forge AGENTS.md is an instructions file published in the GitHub repository srmackey/context-forge (1 stars, last pushed 23d ago), licensed MIT. It adds 2,834 tokens to every session, about $0.0142 per session on Opus 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 2026-08-31.
Other instructions, from other repositories
ai-workflow AGENTS.md
Instructions for cunhaax/ai-workflow, covering ai workflow template, rules — non-negotiable, project overview, commands and architecture.
trackly-cli CLAUDE.md
Claude Code instructions for trackly-app/trackly-cli, covering trackly-cli, tech stack, backend production source of truth, directory structure and key commands.
com-example-config-AsyncTestConfig
Instructions for PIsberg/vibetags, covering copilot instructions for asynctestconfig and immutable type.
pupila CLAUDE.md
Instructions for FranRom/pupila, covering claude.md, where things live, stack, run locally and repo layout.
claude-bridge AGENTS.md
Instructions for kadaba/claude-bridge, covering agents.md — using claude-bridge from any ai coding cli, two independent channels, 1. coordination (the conversation), 2. file transfer (the bytes) — same for every tool and workflows.
UsageAtlas AGENTS.md
Instructions for SRX9/UsageAtlas, covering repository guidelines and coding approach.