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/boardkit/orchestrator/claude-mdgit clone --depth 1 https://github.com/BoardKit/orchestratorWhat 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 | $0.08409 | $0.08409 |
| Opus 5 | $0.04204 | $0.04204 |
| Sonnet 5 | $0.01682 | $0.01682 |
| Haiku 4.5 | $0.00841 | $0.00841 |
Grade D, and why
orchestrator CLAUDE.md scanned grade D with 3 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.
Reads agent configuration directoriesmediumAgent snooping
.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.
- `your-repo/.claude/settings.json -> ../../../orchestrator/shared/settings/your-repo/settings.json` Enumerates other installed skillsmediumAgent snooping
Other skills' SKILL.md files reveal prompts, capabilities and secrets that should be invisible to peers.
ls -la your-repo/.claude/skills/ Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
rm -rf setup/ How it starts
The opening of the file, as written. The whole thing — 990 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Orchestrator - Claude Code Context
⚠️ SETUP REQUIRED - Run /setup-orchestrator to configure this for your organization.
Last Updated: [DATE - will be updated by setup wizard]
Repository Overview
What is this repo? The Orchestrator is a central infrastructure repository that provides shared Claude Code resources (agents, skills, hooks, commands) to all projects via symlinks.
Purpose: Single source of truth for:
- Generic AI agents (code review, planning, documentation, refactoring)
- Auto-triggering skills (organization-specific development patterns)
- Hooks (skill activation, file tracking)
- Slash commands (dev docs creation)
- Shared guidelines (architecture, error handling, testing)
Key Difference from Application Repos:
- orchestrator (this repo): Shared infrastructure provider
- your repositories: [Will be listed here after setup - e.g., app repo, core repo, etc.]
Critical: Changes here affect ALL repos via symlinks. Test in application repos before committing.
Resource Discovery Map
This section lists everything available in the orchestrator. Use this as your reference for what exists and when to use it.
Skills (Auto-Trigger)
Skills activate automatically when you edit files matching specific patterns or keywords.
[REPOSITORY-SPECIFIC SKILLS WILL BE GENERATED DURING SETUP]
After running /setup-orchestrator, you will have one skill per repository, like:
[repo-name]-guidelines (Example)
When it triggers: Editing files in your [repo-name] repo
- File patterns: [e.g., **/repo-name/**/*.tsx, **/repo-name/**/*.py]
- Keywords: [e.g., React, FastAPI, etc. - detected from your tech stack]
What it provides:
- [Tech stack-specific development patterns]
- [Error handling guidance for your frameworks]
- [Testing patterns for your tools]
References: guidelines/error-handling.md, guidelines/testing-standards.md
Status: ⏳ Will be created during setup
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 First seen · 990 lines · 8,409 tokens per session scan D d25069de3b84
orchestrator CLAUDE.md is an instructions file published in the GitHub repository BoardKit/orchestrator (37 stars, last pushed 5mo ago), licensed MIT. It adds 8,409 tokens to every session, about $0.0420 per session on Opus 5. A static security scan graded it D with 3 findings (reads agent configuration directories, enumerates other installed skills, recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other instructions, from other repositories
agent-orchestrator AGENTS.md
AGENTS.md instructions for Untrivial-ai/agent-orchestrator, covering agents.md, repo layout, commands, where to look first and distribution.
agent-orchestrator CLAUDE.md
Claude Code instructions for Untrivial-ai/agent-orchestrator, covering claude.md, app state lives under /.ao only and design system.
DashClaw copilot-instructions.md
Instructions for ucsandman/DashClaw, covering dashclaw — github copilot instructions, design context, users, brand personality and aesthetic direction.
awesome-ai-gateway CLAUDE.md
Instructions for cuihuan/awesome-ai-gateway, covering claude.md — working rules for this repository, what this repository is, the two goals — judge every change against these, non-negotiable rules and before you commit.
forceinjection.github.io AGENTS.md
Instructions for ForceInjection/forceinjection.github.io, covering 项目概览与工具使用说明, 2 jekyll 本地开发环境, ruby 版本管理, 本地启动命令 and 在项目根目录执行（确保 rbenv 已激活 3.2.7）.
holon CLAUDE.md
Instructions for holon-run/holon: This file is read by Claude Code and other agent tooling.