Borrowing it
Nothing to install: this file belongs to ndycode/codex-multi-auth. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/ndycode/codex-multi-auth/main/AGENTS.mdgit clone --depth 1 https://github.com/ndycode/codex-multi-authWrote 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/ndycode/codex-multi-auth/agents-md)<a href="https://agentmods.dev/instructions/ndycode/codex-multi-auth/agents-md"><img src="https://agentmods.dev/badge/instructions/ndycode/codex-multi-auth/agents-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.
<a href="https://agentmods.dev/instructions/ndycode/codex-multi-auth/agents-md"><img src="https://agentmods.dev/badge/instructions/ndycode/codex-multi-auth/agents-md.svg" alt="Reviewed on agentmods" width="80" 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 | $0.06249 | $0.06249 |
| Opus 5.5 | $0.02500 | $0.02500 |
| Sonnet 5.5 | $0.01250 | $0.01250 |
| Haiku 4.5 | $0.00625 | $0.00625 |
Grade B, and why
codex-multi-auth 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 6d 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.
- Official Codex state: `~/.codex/auth.json`, `~/.codex/accounts.json`, `~/.codex/config.toml` (`cli_auth_credentials_store="file"` is enforced top-level; opt out `CODEX_MULTI_AUTH_ENFORCE_CLI_FILE_AUTH_STORE=0`). How it starts
The opening of the file, as written. The whole thing — 210 lines — stays where its author put it; the contents beside it link to each section on GitHub.
PROJECT KNOWLEDGE BASE
Generated: 2026-10-11 Commit: 40c1e2e7 Branch: main Package version: 2.24.0
OVERVIEW
codex-multi-auth is a Codex CLI-first OAuth account manager and optional forwarding wrapper for the official Codex CLI. The architecture is manager-first: the codex-multi-auth bin (scripts/codex-multi-auth.js) owns the account-management CLI — 31 subcommands dispatched through lib/codex-manager.ts (runCodexMultiAuthCli), codex-multi-auth login ≡ codex-multi-auth auth login. The codex-multi-auth-codex bin (scripts/codex.js) owns explicit wrapper forwarding to the official Codex CLI plus shadow CODEX_HOME and runtime-rotation-proxy setup; mcodex is a launcher-only convenience wrapper over scripts/codex.js; codex-multi-auth-app-launcher is a reversible OS launcher routing helper. Runtime rotation is default-on: forwarded Responses/model traffic routes through a loopback-only account-rotation proxy (provider id codex-multi-auth-runtime-proxy). Local governance (usage ledger, budget guards, account policies, routing profiles, quota cache, local bridge tokens) stays file-backed under the multi-auth root. index.ts remains the exported plugin-host compatibility entry, but the primary product surface is the account manager, optional wrapper, storage, runtime proxy, governance commands, and repair tooling.
STRUCTURE
./
├── scripts/
│ ├── codex-multi-auth.js # `codex-multi-auth` bin: standalone account-manager CLI entrypoint
│ ├── codex.js # `codex-multi-auth-codex` bin: official-CLI forwarder, shadow CODEX_HOME + runtime proxy setup
│ ├── mcodex.js # `mcodex` bin: launcher only — forwards to codex.js; optional --monitor / --tmux
│ ├── codex-app-launcher.js # `codex-multi-auth-app-launcher` bin: reversible Windows/macOS/Linux launcher routing
│ ├── codex-routing.js # auth-subcommand routing + compatibility alias normalization
│ ├── codex-bin-resolver.js # official Codex binary discovery (env override → npm → prefix roots → PATH)
│ ├── codex-app-router.js # detached persistent localhost router behind `rotation bind-app` (shipped, not a bin)
│ ├── postinstall.js # notice-only npm hook; real setup is lib/runtime/first-run.ts
│ ├── preuninstall.js # shipped but not an npm hook; `uninstall` subcommand replicates it
│ ├── install-codex-auth-utils.js # installer helpers (install-codex-auth.js is repo-only, not shipped)
│ ├── repo-hygiene.js # `clean`/`check` repo hygiene + Windows retry helpers
│ ├── copy-oauth-success.js, generate-config-schema.mjs # build tooling
│ ├── check-pack-budget.mjs + -lib.js # package size budget check
│ ├── test-model-matrix.js, benchmark-*.mjs, bench-format/, test-all-models.sh, validate-model-map.sh # matrix/bench tooling
│ └── audit-dev-allowlist.js, update-vendor-provenance.mjs, verify-vendor-provenance.mjs # audit/vendor tooling
├── index.ts # plugin-host compatibility entry (`.` export: OpenAIOAuthPlugin)
├── lib/ # core runtime logic (see lib/AGENTS.md)
│ ├── auth/ # OAuth flow, PKCE, dual-bind callback server, device auth, token utils
│ ├── storage.ts + storage/ # V3 account pool: load/save, three-way snapshot merge, dedup, backups/WAL, flagged pool, file/json-store locks
│ ├── accounts.ts + accounts/ # pool facade, health scoring, cooldowns, per-account rate limits
│ ├── runtime-rotation-proxy.ts + runtime/ # loopback Responses/model proxy; app bind, first-run, native binding, selection, quota, checks
│ ├── request/ # request transform, headers, SSE handling, failover/backoff policy
│ ├── policy/ # runtime policy composition + hot-path store caches
│ ├── usage/ # append-only local usage ledger, pricing, redaction
│ ├── codex-cli/ # official Codex CLI state sync + writer helpers
│ ├── codex-manager.ts + codex-manager/ # CLI dispatcher, command modules, settings hub, login menus
│ ├── prompts/ # model-family prompts, GitHub ETag cache
│ ├── context-budget*/ + context-overflow.ts # proactive budget guard + reactive overflow → synthesized replies
│ ├── oc-chatgpt-*.ts # oc-chatgpt multi-auth interop (target detect, import adapter, sync orchestrator)
│ ├── recovery/ # conversation recovery state
│ ├── tools/ # hashline helper tools
│ ├── ui/ # TUI rendering, menus, copy, theme, select
│ └── (flat modules) # governance stores, quota, refresh machinery, rotation, config, schemas, utils
├── test/ # vitest suites (see test/AGENTS.md) — 424 files incl. property/ + chaos/
├── docs/ # user docs + reference/, releases/ (~87), development/ runbooks, audits/, design/, benchmarks/
├── config/ # pluginConfig templates (modern/legacy/minimal) + generated JSON schema
├── vendor/ # vendored codex-ai-plugin + codex-ai-sdk type-only shims (+ provenance.json)
├── assets/ # static assets (icons, hero image)
├── bench/ # format benchmark harness
├── skills/ # codex-auth-setup skill (referenced by plugin manifest; not in files[])
├── .github/ # CI workflows, issue/PR templates, plugin scanner fixtures
└── dist/ # build output (generated, do not edit)
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.
- 6d ago Changed 1f97c98d5961
- 10d ago Changed e08e902a427f
- 12d ago Changed 10f94c699239
- 15d ago Changed · +2 tokens per session c617a923fc74
- 20d ago Changed · +3 tokens per session e729a8390ed4
- 24d ago Changed · -2 tokens per session 8608492b58d7
- 27d ago Changed · +8 tokens per session 63902feb5fb2
- 28d ago Changed · +65 lines · +3,471 tokens per session 872da7f309ff
- 29d ago Changed 124f795a3a68
- 1mo ago First seen · 145 lines · 2,767 tokens per session scan B c11f5c9f5b72
codex-multi-auth AGENTS.md is an instructions file published in the GitHub repository ndycode/codex-multi-auth (556 stars, last pushed 7d ago), licensed MIT. It adds 6,249 tokens to every session, about $0.0250 per session on Opus 5.5. 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 2026-08-30.
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).
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.
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).
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.
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).
deno copilot-instructions.md
Copilot instructions for denoland/deno, covering deno development guide for github copilot, network access, high level overview, key directories and key files to understand first.