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/borda/ai-rig/claude-mdgit clone --depth 1 https://github.com/Borda/AI-RigWrote 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.
[](https://agentmods.dev/instructions/borda/ai-rig/claude-md)<a href="https://agentmods.dev/instructions/borda/ai-rig/claude-md"><img src="https://agentmods.dev/badge/instructions/borda/ai-rig/claude-md.svg" alt="Measured on agentmods" height="20"></a>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 | $0.02000 | $0.02000 |
| Opus 5 | $0.01000 | $0.01000 |
| Sonnet 5 | $0.00400 | $0.00400 |
| Haiku 4.5 | $0.00200 | $0.00200 |
Grade A, and why
AI-Rig CLAUDE.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 2d ago.
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 — 120 lines — stays where its author put it; the contents beside it link to each section on GitHub.
- Any policy change in one listed instruction file must trigger a relevance review of every other listed file before completion.
- Synchronize applicable shared policy in either direction.
- Preserve intentional agent-specific differences.
- Record when no counterpart change is needed.
Instruction Layering
Repository-wide policy belongs in this top-level file. Lower-scope instruction files inherit it and must add only narrower rules or explicit exceptions, never repeat the same policy.
When a top-level policy changes, review lower layers for conflicts or obsolete duplication rather than copying the new text into them.
Edit Scope — Hard Constraint
ALL edits must stay within this project directory. Never directly edit ~/.claude/ or $CODEX_HOME (cache, settings, hooks, or any file under the user home).
Permitted edit roots (project-local):
.claude/settings.json,.claude/settings.local.json— project Claude config.codex/— project Codex config, skills, session policy (mirrored to$CODEX_HOMEbyMakefile, never edited there)plugins/*/{agents,skills,rules,hooks,bin}/— plugin source
Propagation to live cache at ~/.claude/plugins/cache/:
Makefileinstalls from the pushed GitHub remote, not local working tree — commit and push first, thenmake sync-claude- Never run
make sync-*/make clear-*against uncommitted/unpushed changes — cache will not reflect them - Never suggest or initiate propagation mid-workflow
- Applies to all skills — no skill auto-syncs
Lint/Format — Use pre-commit Hooks, Not Direct Tools
Repo pins lint/format tools via .pre-commit-config.yaml (ruff, eslint, mdformat, prettier, codespell, etc). Never invoke these tools directly (ruff check, ruff format, eslint, mdformat, ...) — version/config drift vs CI.
Invoke the specific hook instead:
pre-commit run <hook-id> --files <path> # single hook, targeted files
pre-commit run --all-files # full sweep
pre-commit run <hook-id> --all-files # single hook, repo-wide
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.
- 2d ago Changed · +4 lines · +48 tokens per session ae0990d72846
- 5d ago First seen · 116 lines · 1,952 tokens per session scan A 838cdfeddb4e
AI-Rig CLAUDE.md is an instructions file published in the GitHub repository Borda/AI-Rig (26 stars, last pushed yesterday), licensed Apache-2.0. It adds 2,000 tokens to every session, about $0.0100 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 2026-08-30.
Other instructions, from other repositories
Claude-Code-Agent-Monitor CLAUDE.md
Claude Code instructions for hoangsonww/Claude-Code-Agent-Monitor, covering claude code working guide, project mission, repo map, non-negotiable engineering rules and commands you should know.
Claude-Code-Agent-Monitor AGENTS.md
AGENTS.md instructions for hoangsonww/Claude-Code-Agent-Monitor, covering codex project instructions, project intent, priorities, where to work and validation expectations.
skills AGENTS.md
AGENTS.md instructions for huangwb8/skills, covering skills - 项目指令, 目录结构与边界, python 与工具运行产物归档, ai 任务中间文件与 skill 协作工作区 and 任务根目录与生命周期.
claude CLAUDE.md
Claude Code instructions for Caseous-kerrcell366/claude, covering @aiai-go/claude — project handbook, project identity, brand rules, tech stack and architecture.
.agents AGENTS.md
AGENTS.md instructions for Weaverse/.agents, covering agent guidelines for .agents, repository structure, build, lint & test commands, validation commands and validate json configuration.
init-claude-md CLAUDE.md
Claude Code instructions for LetheChen/init-claude-md, covering agents.md, session maintenance protocol, core rules (from karpathy), language and project conventions.