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/nullorder/agenthub/agents-mdgit clone --depth 1 https://github.com/nullorder/agenthubWhat 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 | $0.01233 | $0.01233 |
| Opus 5 | $0.00616 | $0.00616 |
| Sonnet 5 | $0.00247 | $0.00247 |
| Haiku 4.5 | $0.00123 | $0.00123 |
Grade A, and why
agenthub 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 — 137 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AgentHub — Agent Guidelines
This file is for AI agents (Claude Code, Copilot, etc.) working on this repository. It describes the project, its conventions, and how to perform common tasks correctly.
What is this project?
AgentHub is a community-driven plugin marketplace for Claude Code. It is a catalog only — plugin code lives in contributors' own repositories (or npm packages). This repo maintains a registry of plugin references in .claude-plugin/marketplace.json.
Project structure
agenthub/
├── plugins/ # One JSON file per plugin (the source of truth)
├── scripts/
│ └── build-marketplace.sh # Aggregates plugins/*.json into .claude-plugin/marketplace.json
├── .github/
│ ├── workflows/
│ │ ├── build-marketplace.yml # Rebuilds .claude-plugin/marketplace.json on push to main
│ │ └── validate-plugin.yml # Validates plugin JSON on PRs
│ └── PULL_REQUEST_TEMPLATE/
│ ├── add-plugin.md
│ ├── update-plugin.md
│ └── remove-plugin.md
├── .claude-plugin/marketplace.json # AUTO-GENERATED — never edit directly
├── CONTRIBUTING.md # Contribution guidelines for humans
├── README.md # User-facing install & usage docs
└── LICENSE # MIT
Critical rules
-
Never edit
.claude-plugin/marketplace.jsondirectly. It is auto-generated byscripts/build-marketplace.shfrom the files inplugins/. If you need to change the marketplace output, modify the plugin files or the build script. -
One plugin = one file. Each plugin is a standalone JSON file in
plugins/named<plugin-name>.json. The filename must match thenamefield inside the JSON. -
Plugin naming conventions:
- Kebab-case only:
my-cool-plugin, notMyCoolPlugin - Be descriptive:
react-test-generator, notrtg - Never use "claude", "anthropic", or "official" in plugin names
- Kebab-case only:
-
Valid categories:
development,testing,devops,security,documentation,productivity,data,design,other
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 · 137 lines · 1,233 tokens per session scan A e3434727d1bf
agenthub AGENTS.md is an instructions file published in the GitHub repository nullorder/agenthub (3 stars, last pushed 2d ago), licensed MIT. It adds 1,233 tokens to every session, about $0.0062 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
agent-powerups CLAUDE.md
Instructions for yeaight7/agent-powerups, covering claude.md, commands, architecture, catalog.json — single source of truth and cli command routing (src/cli/apx.ts).
agent-powerups GEMINI.md
Instructions for yeaight7/agent-powerups, covering project overview, main technologies, architecture/structure, building and running and development conventions.
pnCore AGENTS.md
AGENTS.md instructions for perniemann/pnCore, covering agents.md, learned user preferences and learned workspace facts.
claude-code-plugins AGENTS.md
Instructions for melodic-software/claude-code-plugins, covering claude-code-plugins, validate a change and conventions that load on demand.
codex AGENTS.md
AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.
buildNext
Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).