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/runnithan/claudex-setup/agents-mdgit clone --depth 1 https://github.com/runnithan/claudex-setupWhat 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 | $0.00706 | $0.00706 |
| Opus 5 | $0.00353 | $0.00353 |
| Sonnet 5 | $0.00141 | $0.00141 |
| Haiku 4.5 | $0.00071 | $0.00071 |
Grade B, and why
claudex-setup AGENTS.md scanned grade B with 1 finding 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.
`~/.codex/config.toml`. Adding a new top-level file? It must also be added to the `ALLOWED` How it starts
The opening of the file, as written. The whole thing — 41 lines — stays where its author put it; the contents beside it link to each section on GitHub.
claudex-setup: Codex guide
Codex reads this file the way Claude Code reads CLAUDE.md. This repo is the claudex hub
(Claude Code + Codex): you run it from here to mine agentic-coding lessons and to
audit your other projects' agent setups in place.
CLAUDE.md is authoritative, read it first
The working rules are shared between both tools and live in CLAUDE.md. Read it
before touching anything. Architecture, the transcript→lessons pipeline, Conventional Commits and
the no-Co-Authored-By rule, CHANGELOG discipline, and the validation steps all live there and
apply to Codex verbatim. They are deliberately not restated here, because a second copy would drift,
which is exactly what this repo is built to prevent. If a convention changes, it changes in
CLAUDE.md only.
The short safety net below is the one deliberate exception. Codex reads AGENTS.md but does not
automatically read CLAUDE.md, so the rules where a miss is catastrophic and unrecoverable are
mirrored here, and only those, because they effectively never change. If this file and
CLAUDE.md ever disagree, CLAUDE.md wins.
- Never commit secrets.
.envfiles, private keys, and service-account JSON stay gitignored. - Public mirror + leak-gate. This private repo publishes a sanitized public mirror via
scripts/publish.sh, which stripsprojects/+transcripts/, freezes the public lesson set, and hard-fails on any leak. Runbash scripts/publish.sh(dry run) before publishing. - Never name a private project in
CHANGELOG.md: it is not in the scrub set, so a private name written there fails the leak gate. Describe the feature, not the project.
Codex-specific notes
- This repo's Codex config surface is exactly what
/optimise --codexinventories anywhere else: thisAGENTS.md, a project.codex/directory (none here yet), and the relevant~/.codex/config.toml. Adding a new top-level file? It must also be added to theALLOWEDpath allowlist inscripts/publish.sh, or publishing hard-fails by design. - Tracked Codex usage lives in the registry: a multi-tool project gets a
projects/<slug>/codex/sub-area (its owncurrent.md/improvements.md/applied-improvements.md/habits.md), audited by/optimise <slug> --codexand keyed<slug>/codexinprojects/audit-log.md. - Honest scope. The lessons library is Claude-Code-first:
/extract-social-lessonsmines a Codex/OpenAI voice cluster, but keeps only lessons that transfer to general agentic coding. Codex is first-class in what this repo audits and tracks, not in what it ships. The slash commands and hooks are Claude Code artifacts and run there.
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 · 41 lines · 706 tokens per session scan B f825bc7cf440
claudex-setup AGENTS.md is an instructions file published in the GitHub repository runnithan/claudex-setup (1 stars, last pushed 2d ago), licensed MIT. It adds 706 tokens to every session, about $0.0035 per session on Opus 5. A static security scan graded it B with 1 finding (reads agent configuration directories). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other instructions, from other repositories
langgraph AGENTS.md
AGENTS.md instructions for langchain-ai/langgraph, covering agents instructions, corridor security analysis, libraries and dependency map.
omnigent copilot-instructions.md
Copilot instructions for omnigent-ai/omnigent, covering copilot code review instructions, e2e test requirement, backend test coverage and frontend test coverage.
AutoAgents AGENTS.md
Instructions for liquidos-ai/AutoAgents, covering autoagents repository guidelines, project structure & module organization, code quality, naming conventions and error handling.
claude-code-karma CLAUDE.md
Instructions for JayantDevkar/claude-code-karma, covering claude.md, project overview, quick start, clone (no submodule init needed — this is a monorepo) and start api (terminal 1).
hatch3r CLAUDE.md
Instructions for hatch3r/hatch3r, covering hatch3r — development instructions, architecture, development commands, two-axis pillar framework (2.0.0) and orchestrator self-discipline (bypass protection).
aeon CLAUDE.md
Instructions for aeonfun/aeon, covering aeon, how aeon works, strategy, voice and soul file hierarchy (read in this order).