moai-adk AGENTS.md

moai-adk AGENTS.md is an instructions file for Codex, OpenCode from modu-ai/moai-adk. It costs 3,222 tokens per session, scanned A, original, Apache-2.0.

A repository policy file that tells coding agents how to work in the MoAI-ADK project. It covers evidence for claims, verification, Git branches, shared checkouts, and worktrees.

In plain words
What is it for?
Use it as the standing guide for making, checking, and reporting code changes in this repository.
Why use it?
It helps prevent agents from claiming that work is complete or tested without evidence and clarifies how changes should be handled in Git.

Instructions file for CodexOpenCode

About the project

MoAI-ADK is a Go-based harness that organizes and verifies Claude Code work across planning, implementation, synchronization, and review stages. Developers use it to structure agentic coding tasks, apply quality gates, and route work across language models, while the catalogue entries extend its workflow with skills, hooks, commands, MCP servers, instructions, and settings.

modu-ai/moai-adk · 1,200 stars · on GitHub

Install

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.

agentmods
npx agentmods add instructions/modu-ai/moai-adk/agents-md
Clone the repo
git clone --depth 1 https://github.com/modu-ai/moai-adk

Made for: Codex, OpenCode.

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 moai-adk AGENTS.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/modu-ai/moai-adk/agents-md.svg)](https://agentmods.dev/instructions/modu-ai/moai-adk/agents-md)
Your own site
<a href="https://agentmods.dev/instructions/modu-ai/moai-adk/agents-md"><img src="https://agentmods.dev/badge/instructions/modu-ai/moai-adk/agents-md.svg" alt="Measured on agentmods" height="20"></a>
Per session 3,222 This file is loaded in full into every session.
When invoked 3,222 The same file — it is already loaded in full.
Security scan A 0 findings. Scan, not verified.
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 $0.03222 $0.03222
Opus 5 $0.01611 $0.01611
Sonnet 5 $0.00644 $0.00644
Haiku 4.5 $0.00322 $0.00322

Measured yesterday against content hash eda0882052e3, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

moai-adk 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 yesterday.

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.

Origin

Copies of this mod

1 near-identical copy found in the catalogue:

AGENTS.md · 243 lines

How it starts

The opening of the file, as written. The whole thing — 243 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md — standing contract for agents in this repository

Every clause here binds a turn regardless of which agent harness drives it. The file is self-sufficient: it assumes no other instruction file is loaded, and no nested AGENTS.md exists anywhere in this repository.

Budget warning. A personal ~/.codex/AGENTS.md joins the same merged chain and is consumed before this file, narrowing what the project's contract can carry. Overflow is dropped from the tail, silently — no warning, no stderr, exit 0. Clauses below are ordered most-critical-first for that reason.

Obligations are carried from .claude/rules/moai/** and CLAUDE.md, which remain the source of truth; compression removed rationale and incident records, never an obligation. Claude-only mechanisms (the question channel, subagent spawning, skills, session handoff) stay there.


1. Evidence and verification claims

No unobserved claim. An actor MUST NOT assert a verification, a completion, a defect / debt / drift, OR the premise underlying a recommendation it did not actually verify with the domain's mechanical tooling. Evidence absent is not evidence of success — nor of failure. The absence of a failure signal never establishes that a check passed; a text-pattern inference is a hypothesis, not a verified defect; a reference existing does not establish that the referenced capability is still live. Reachability is not justification.

Baseline-integrity attribution. Every verification claim MUST be attributed to an actually-measured baseline — the command that was run plus the output observed, in this run, against this tree. A figure carried over from another package, tree, or point in time is not a baseline; using it as a fresh measurement violates this. Anything unattributed is a Gap, not a Claim.

Evidence-bearing report format. Verification and completion reports SHOULD carry five sections, on every report and not only the first: Claim (what is asserted); Evidence (the command run plus its verbatim output — a summary is not evidence); Baseline-attribution (what it was measured against, in this run); Gaps (what was explicitly NOT observed — an empty Gaps section asserts nothing was left unobserved, which must itself be true); Residual-risk (what could still be wrong despite what was observed).

Read the full file on GitHub · 243 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. yesterday First seen · 243 lines · 3,222 tokens per session scan A eda0882052e3

Subscribe to this mod's changes

moai-adk AGENTS.md is an instructions file published in the GitHub repository modu-ai/moai-adk (1,200 stars, last pushed 2d ago), licensed Apache-2.0. It adds 3,222 tokens to every session, about $0.0161 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 2026-09-03.