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/kensaurus/mushi-mushi/agents-mdgit clone --depth 1 https://github.com/kensaurus/mushi-mushiWrote 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/kensaurus/mushi-mushi/agents-md)<a href="https://agentmods.dev/instructions/kensaurus/mushi-mushi/agents-md"><img src="https://agentmods.dev/badge/instructions/kensaurus/mushi-mushi/agents-md.svg" alt="Measured on agentmods" 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 | $0.12588 | $0.12588 |
| Opus 5 | $0.06294 | $0.06294 |
| Sonnet 5 | $0.02518 | $0.02518 |
| Haiku 4.5 | $0.01259 | $0.01259 |
Grade A, and why
mushi-mushi 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 4d 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 — 853 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md — Mushi Mushi Agent Conventions
This file describes the autonomous agents, edge functions, and orchestration patterns used across the Mushi Mushi monorepo. It follows the OpenAI ExecPlans philosophy: self-contained, machine-readable plans that any AI coding agent can execute without additional context.
Positioning — read this before you touch any user-facing surface
The canonical constitution is
/VISION.md. This block is the compressed copy so on-message behavior survives even when an agent only loadsAGENTS.md. If this andVISION.mdever disagree,VISION.mdwins — fix it here, don't fork it. Drift is enforced against byscripts/check-positioning-consistency.mjs.
North star. Your AI shipped it. Mushi tells you why it broke — in plain English, in your editor, with the fix ready to go — so a bug costs you five minutes instead of your whole afternoon.
Category we own. The bug mediator for AI-built apps — one queue between your users, your monitoring (Sentry/Crashlytics/Rollbar), your tracker (Linear/Jira/GitHub), your chat (Slack/Discord/Teams), and your coding agents, where every bug gets a plain-English diagnosis and a ready fix. (Not "error monitoring", not "observability" — those are someone else's ops-coded words.)
Primary buyer. The solo / indie vibe coder who builds fast with AI (Cursor, Claude Code, Lovable, Bolt), ships to real users, then loses afternoons when something breaks because they don't fully grasp the generated code. Small teams and agencies are secondary; the enterprise SRE running Sentry + Datadog + Firebase is explicitly not who we lead with.
The three things we will not do (drift tripwires):
- We will not require a monitoring stack to get value. Standalone-first, always — every inbound integration is an optional on-ramp, never a prerequisite.
- We will not make you rip anything out. Sentry, Linear, Slack, your agents all stay — Mushi mediates between them. Enterprise plumbing (SSO, region routing, Helm) still lives in
docs/operators/and never leads. - We will not let the surfaces diverge. The tagline, the north-star sentence, and the buyer stay identical across npm, repo, landing, and this file.
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.
- 4d ago First seen · 853 lines · 12,588 tokens per session scan A 30d6beb53a8f
mushi-mushi AGENTS.md is an instructions file published in the GitHub repository kensaurus/mushi-mushi (3 stars, last pushed 7d ago), licensed MIT. It adds 12,588 tokens to every session, about $0.0629 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 2026-08-31.
Other instructions, from other repositories
Tracely-ai CLAUDE.md
Claude Code instructions for Jwuthri/Tracely-ai, covering claude.md, commands, architecture, hard rules and gotchas.
fundamental-ngx AGENTS.md
AGENTS.md instructions for SAP/fundamental-ngx, covering angular 22+ development guidelines, persona, quick reference, essential commands and build/lint/test specific library.
primitives CLAUDE.md
Claude Code instructions for radix-ng/primitives, covering radix ng primitives — claude guide, project knowledge, project overview, monorepo structure and naming conventions.
openbridge-webcomponents watch-radial-instruments.instructions.md
Instructions for Ocean-Industries-Concept-Lab/openbridge-webcomponents, covering watch & radial instruments, architecture overview, key principle: logic in watch.ts, instruments stay thin, shared sibling modules and layering & stacking pattern.
openbridge-webcomponents jsdoc.instructions.md
Instructions for Ocean-Industries-Concept-Lab/openbridge-webcomponents, covering jsdoc documentation rules, slots and events are consumer-critical (@slot / @fires), conditional properties (@availablewhen), what the tag does in storybook and comment style.
openbridge-webcomponents line-area-charts.instructions.md
Instructions for Ocean-Industries-Concept-Lab/openbridge-webcomponents, covering line/area charts & gauge trend, architecture overview, key principle: logic in base class, subclasses stay thin, base class architecture and line graph (non-filled).