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/mibayy/token-savior/agents-mdgit clone --depth 1 https://github.com/Mibayy/token-saviorWhat 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 | $0.01890 | $0.01890 |
| Opus 5 | $0.00945 | $0.00945 |
| Sonnet 5 | $0.00378 | $0.00378 |
| Haiku 4.5 | $0.00189 | $0.00189 |
Grade A, and why
token-savior 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 3d 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.
How it starts
The opening of the file, as written. The whole thing — 146 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Token Savior — configuration recommandée, tous agents
Ce fichier suit la convention AGENTS.md, lue par une trentaine d'agents
(OpenAI Codex, Cursor, GitHub Copilot, Gemini CLI, Aider, Zed, Windsurf,
Jules, Factory, Devin…). Claude Code lit CLAUDE.md, dont le contenu est
équivalent. Déposez ce fichier à la racine de votre projet pour reproduire
le résultat de tsbench : 97,9 % du score,
−80 % de tokens actifs.
Ce qui est portable, et ce qui ne l'est pas
| Couche | Portabilité | Détail |
|---|---|---|
| Serveur MCP (69 outils) | Universelle | MCP est un protocole ouvert. Aucune adaptation. |
AGENTS.md |
~30 agents | Ce fichier. |
Skills (SKILL.md) |
16 à 32 outils | Standard ouvert agentskills.io. |
| Hooks | Aucun standard | Chaque agent a son format. C'est la seule couche à écrire. |
ts init --agent {claude,cursor,gemini,codex,openclaw} écrit la couche hooks
pour vous, au bon endroit et au bon format.
Les trois modèles de hooks rencontrés
Il n'y a pas un format à traduire, il y a trois architectures différentes.
| Agent | Modèle | Emplacement |
|---|---|---|
| Claude Code, Codex | Commande shell par événement | settings.json / hooks.json |
| OpenClaw | Hook pack : dossier HOOK.md + handler.js |
openclaw.json → hooks.internal.load.extraDirs |
| Hermes | Aucun système de hooks connu | MCP seul, ~/.hermes/config.yaml |
OpenClaw n'expose aucun événement d'outil : ses prédicats sont
isAgentBootstrapEvent, isGatewayStartupEvent, isMessageReceivedEvent,
isMessagePreprocessedEvent, isMessageSentEvent, isMessageTranscribedEvent
et isSessionPatchEvent. Ni tool_capture ni le réécriveur Bash ne peuvent y
fonctionner. Le moteur mémoire, lui, porte : voir
hooks/openclaw/token-savior-memory/.
Son injection de contexte ne passe pas par la sortie standard mais par la
mutation de context.bootstrapFiles pendant agent:bootstrap, avec des
entrées de forme { name, path, content, missing }. Le consommateur
déduplique par nom de fichier : ne jamais injecter sous AGENTS.md, cela
écraserait le vrai fichier de l'agent.
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.
- 3d ago First seen · 146 lines · 1,890 tokens per session scan A 78bd57d40f44
token-savior AGENTS.md is an instructions file published in the GitHub repository Mibayy/token-savior (1,137 stars, last pushed 23d ago), licensed MIT. It adds 1,890 tokens to every session, about $0.0095 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 2026-08-30.
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).
spec-kit AGENTS.md
AGENTS.md instructions for github/spec-kit, covering agents.md, about spec kit and specify, quickstart — add a new integration in 5 steps, integration architecture and integrationmanifest — file tracking.
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.
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 oss-third-party-notices.instructions.md
Instructions for microsoft/vscode, covering vs code oss third-party-notices pipeline, architecture, pipeline flow in ci, applying the notice (cutover) and fallback chain (never fail the build).
next.js AGENTS.md
Instructions for vercel/next.js, covering next.js development guide, codebase structure, monorepo overview, core package: packages/next and other important packages.