agentmods.dev

claudex-stereo: Instructions file for Claude Code

CLAUDE.md

claudex-stereo CLAUDE.md is an instructions file for Claude Code, Codex from vsladkov/claudex-stereo. It costs 5,702 tokens per session, scanned A, original, Apache-2.0.

Contributor instructions for claudex-stereo, a plugin that pairs Claude with the OpenAI Codex command-line tool so one can build while the other reviews and checks the work.

In plain words
What is it for?
Use them when developing or reviewing the plugin, preserving its directory structure, adding provider models, running its checks, or preparing an installable package.
Why use it?
They explain the repository layout, dependency rules, development loop, tests, and packaging boundary that contributors must follow.

Instructions file for Claude CodeCodex

Written for Claude Code and Codex: $ARGUMENTS substitution, but also reads $CODEX_HOME. Also seen: reads .claude/ paths; mentions .claude-plugin; mentions CLAUDE.md.

This is vsladkov/claudex-stereo's own configuration. It tells Claude Code and Codex how to work on claudex-stereo itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything claudex-stereo configures →

not rated 5repo +1 changed 29d ago A scan 5,702 tokens original Apache-2.0
View source ↗ vsladkov/claudex-stereo claudex-stereo.com
Reuse

Borrowing it

Nothing to install: this file belongs to vsladkov/claudex-stereo. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/vsladkov/claudex-stereo/main/CLAUDE.md
Clone the repo
git clone --depth 1 https://github.com/vsladkov/claudex-stereo

Made for: Claude Code, Codex.

Wrote 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.

agentmods badge for claudex-stereo CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/vsladkov/claudex-stereo/claude-md/github.svg)](https://agentmods.dev/instructions/vsladkov/claudex-stereo/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/vsladkov/claudex-stereo/claude-md"><img src="https://agentmods.dev/badge/instructions/vsladkov/claudex-stereo/claude-md/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for claudex-stereo CLAUDE.md

Your own site · 80×15
<a href="https://agentmods.dev/instructions/vsladkov/claudex-stereo/claude-md"><img src="https://agentmods.dev/badge/instructions/vsladkov/claudex-stereo/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 5,702 This file is loaded in full into every session.
When invoked 5,702 The same file — it is already loaded in full.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.05702 $0.05702
Opus 5.5 $0.02281 $0.02281
Sonnet 5.5 $0.01140 $0.01140
Haiku 4.5 $0.00570 $0.00570

Measured 29d ago against content hash 1c4c2dc2ab6e, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

claudex-stereo 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 29d 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.

CLAUDE.md · 340 lines

How it starts

The opening of the file, as written. The whole thing — 340 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Claudex Stereo — contributor notes

Claude Code plugin pairing Claude with the OpenAI Codex CLI: Claude plans and builds, Codex challenges the plan and gates the diff (/stereo:* commands). Published at github.com/vsladkov/claudex-stereo with the site at claudex-stereo.com; development is dogfooded — the plugin under development is also the tool in use.

Layout

  • plugins/stereo/ — the plugin; only this directory ships into installs. That nesting is the packaging boundary (installs copy this dir's working tree): never flatten it, or node_modules/, tests/, and .generated/ ship into every install.
    • scripts/ — thin entry points: codex-companion.ts (the CLI the command markdown invokes), app-server-broker.ts, and the hook entry points session-lifecycle-hook.ts / stop-review-gate-hook.ts, which hooks/hooks.json reaches through the .cjs shims of the same name (session-lifecycle-hook.cjs, stop-review-gate-hook.cjs: each runs node-version-guard.cjs, then imports its .ts). Paths and filenames are load-bearing.
    • src/ — layered, imports point strictly downward: shared → platform → protocol/broker/workspace → transport → models → runtime → jobs → render/workflows → cli/hooks (models, runtime, and jobs are one tier in spirit, but imports within it point that way). protocol/ bridges the codegen types and holds the broker RPC/endpoint constants shared across layers. broker/server.ts is a process entry point whose transport client is injected by scripts/app-server-broker.ts, so broker/ never imports upward. transport/claude-cli.ts spawns claude -p and parses its stream-json events; runtime/claude-runner.ts turns a role request into flags, environment, and a turn result.
    • Models: models/model-table.ts is the one table of model versions for both runtimes (one row per version, with its default effort and, where it lacks a tier, the efforts it takes); its Claude rows are the models Claude Code offers, and adding a row is the only maintenance for a new version. Claude Code has no command that lists its models, and its catalog cache under ~/.claude/cache/ is as private as Codex's: never read it. ROLE_DEFINITIONS in models/role-defaults.ts holds each role's built-in default, which moves only when that table changes. models/registry.ts holds the third-party provider aliases (one row per model). OpenAI models are not rows: models/catalog.ts resolves codex:<family>[-<version>] against the live Codex catalog (model/list, cached at $CODEX_HOME/companion-state/codex-models.json, with a one-model built-in snapshot as the only fallback). Never read Codex's own models_cache.json: it is version-gated per client and lags the live list. models/claude-models.ts holds the claude:<family>[-<version>] grammar; never pass Claude Code a bare alias. Selection, model, and effort resolution live in one launch block (cli/launch.ts, over models/role-defaults.ts) shared by task, plan-review, and the reviews; config and setup render role defaults through it, and the pair commands read each role's dry run instead of re-deriving it. The user-facing rules live in skills/model-routing/SKILL.md.
    • commands/ skills/ agents/ roles/ prompts/ schemas/ hooks/ — the plugin surface Claude actually reads; prompts/schemas are loaded via shared/paths.ts. agents/ holds only codex-rescue.md (the one Agent type, launched by /stereo:rescue); the six Claude role definitions live in roles/, so Claude Code never registers them as Agent types.
  • .claude-plugin/marketplace.json — the repo doubles as the marketplace (claudex-stereo, added by local path or GitHub slug); install source points at ./plugins/stereo.
  • AGENTS.md — the Codex-side contributor entry point; it defers to this canonical CLAUDE.md.
  • .generated/ — gitignored codegen output (type-only imports; regenerated by npm run typecheck's prestep, which needs the codex CLI on PATH).
  • docs/ — the GitHub Pages source for claudex-stereo.com (marketing site, social card, CNAME/robots/sitemap/404); publishing-only.
  • scripts/bump-version.ts, scripts/provider-probe.ts, scripts/test-windows.ts, tests/, docs/, and .github/ — dev-only, never shipped. The provider probe backs npm run provider-probe; the Windows test runner backs npm run test:windows (every test file except its exclusion list).

Read the full file on GitHub · 340 lines

Changes

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.

  1. 29d ago Changed · +174 lines · +3,237 tokens per session 1c4c2dc2ab6e
  2. 1mo ago First seen · 166 lines · 2,465 tokens per session scan A 4b79c37a5fa3

Subscribe to this mod's changes

claudex-stereo CLAUDE.md is an instructions file published in the GitHub repository vsladkov/claudex-stereo (5 stars, last pushed 29d ago), licensed Apache-2.0. It adds 5,702 tokens to every session, about $0.0228 per session on Opus 5.5. 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.

Building agentsClaude CodeCode reviewai-pair-programmingclaudeclaude-code-plugincli
Related

Other instructions, from other repositories

vscode buildNext.instructions.md

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).

microsoft/vscode · 9,715 tokens

next.js AGENTS.md

AGENTS.md instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.

vercel/next.js · 7,461 tokens

vscode oss-third-party-notices.instructions.md

Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, cache eligibility and rollout and applying the notice (cutover).

microsoft/vscode · 5,262 tokens

langchain AGENTS.md

AGENTS.md instructions for langchain-ai/langchain, covering global development guidelines for the langchain monorepo, corridor security analysis, project architecture and context, monorepo structure and development tools & commands.

langchain-ai/langchain · 4,680 tokens

vscode copilot-instructions.md

Copilot instructions for microsoft/vscode, covering vs code copilot instructions, project overview, root folders, core architecture (src/ folder) and built-in extensions (extensions/ folder).

microsoft/vscode · 2,750 tokens

deepseek-harness AGENTS.md

AGENTS.md instructions for deepseek-ai/deepseek-harness, covering agents.md, pre-stable apis and released session data, repository layout, commands and host sandbox failures.

deepseek-ai/deepseek-harness · 4,261 tokens