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/haoruilee/oh-my-mcode/agents-mdgit clone --depth 1 https://github.com/haoruilee/oh-my-mcodeWhat 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 | $0.04160 | $0.04160 |
| Opus 5 | $0.02080 | $0.02080 |
| Sonnet 5 | $0.00832 | $0.00832 |
| Haiku 4.5 | $0.00416 | $0.00416 |
Grade A, and why
oh-my-mcode 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 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.
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 — 180 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
IMPORTANT: This file is the source of truth for agent instructions. Update it (or a nested AGENTS.md) when adding guidance. Do not add a parallel CLAUDE.md / Cursor rules dump.
Commands
| Gate | Command |
|---|---|
| hermetic | npm test |
| hermetic eval | npm run eval (fixture harness; not live mcode) |
| CI | .github/workflows/hermetic.yml runs the two hermetic gates on push/PR to main. No live mcode, no MiniMax secrets. |
| live rematch | oh-my-mcode plan or oh-my-mcode max on a copy of test/fixtures/hello-pkg |
| follow-goal eval | evals/tasks/follow-goal + npm run eval (instruction-follow fixture; not a production ΔY) |
This repo uses npm. Do not use bun or pnpm — those rewrite the lockfile. Do not run a production publish or npm publish in an agent session.
Contributor operating manual for oh-my-mcode. Not a skill-pack blurb. This file does not become the orchestrator.
North star: Not more agents. Ship with evidence. The TypeScript harness owns the loop. mcode is the host. Durable goal_state on the run store and src/guard.ts loop hygiene reconstruct why we stopped (repeat-finding / repair-cap / host-crash). Host timedOut / signal are orthogonal to exitCode. We are not DeepSeek Harness.
Default context
This repo is a verified-delivery harness that drives mcode exec. The plugin / TUI is a window. src/orchestrator.ts owns verify, repair, resume, and evidence. CLI and MCP share one core (src/harness.ts submit).
Unless the user says otherwise, assume work is this package — not a new host product, not a second runtime.
Terminology — when the user says agent, they mean this product (the MiniMax host process + our harness), not you (the assistant in this session).
| Term | Means |
|---|---|
| max | Hero command oh-my-mcode max (alias omm). Owns DISCOVER → PLAN → PLAN_REVIEW → EXECUTE → VERIFY → ACCEPT. |
| plan | oh-my-mcode plan. Stops at PLAN_REVIEW. Does not Accept. |
| worker | One role contract in agents/*.md for the same MiniMax host process. One mcode exec. No grandchildren. |
| mcode | Host binary (@minimax-ai/code). We do not wrap it as mmx / mavis. |
| run store | <workspace>/.minimax/runs/<run_id>/ — run.json, plan.md, tasks.json, events.jsonl, evidence/, findings.json, summary.md. |
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 · 180 lines · 4,160 tokens per session scan A 093934025485
oh-my-mcode AGENTS.md is an instructions file published in the GitHub repository haoruilee/oh-my-mcode (51 stars, last pushed 4d ago), licensed MIT. It adds 4,160 tokens to every session, about $0.0208 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
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.
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).
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.