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/honeycomb-technologies/mitsuro/agents-mdgit clone --depth 1 https://github.com/honeycomb-Technologies/MitsuroWrote 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/honeycomb-technologies/mitsuro/agents-md)<a href="https://agentmods.dev/instructions/honeycomb-technologies/mitsuro/agents-md"><img src="https://agentmods.dev/badge/instructions/honeycomb-technologies/mitsuro/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 | $0.06346 | $0.06346 |
| Opus 5 | $0.03173 | $0.03173 |
| Sonnet 5 | $0.01269 | $0.01269 |
| Haiku 4.5 | $0.00635 | $0.00635 |
Grade A, and why
Mitsuro 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 — 338 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS Guide: Mitsuro
Purpose
Repository-level engineering, product, validation, and release guardrails for Mitsuro: an AI coding assistant with mobile, web, desktop, CLI/TUI, ACP editor integration, and autonomous agent modes.
AGENTS Strategy
This is the only AGENTS file in the repository. All module-specific invariants live in sections below rather than scattered across subdirectories.
Product Language and Compatibility
- Product name: Mitsuro.
- Company name: Honeycomb Technologies.
- Interactive assistant/session: Agent. User-facing modes may be named Chat and Code.
- Durable autonomous system: Hive. An individual delegated worker is a Hive Worker.
- Hive surfaces: Workers, Groups, Activity, Calendar, and Memory.
- Activity accent/state: Pulse, only where the product design calls for that term.
- Public product copy, screenshots, package descriptions, release notes, and repository prose should use Mitsuro, Agent, Hive, Hive Worker, Hive Workers, Groups, Activity, Calendar, Memory, and Pulse consistently.
- The canonical public repository is
honeycomb-Technologies/Mitsuro. Mobile launch URLs usemitsuro://. - Canonical identifiers include
mitsuro,mitsuro-*,mitsuro-hive,@mitsuro/*,/api/hive/*,session_type = "hive",~/.mitsuro, and the corresponding Expo, native, database, and deployment names. - Prior identifiers may appear only in dedicated, tested compatibility readers and migrations. Those boundaries read prior state and write canonical state; they are not current product language.
- Any compatibility retirement requires an explicit plan covering stored data, installed clients, deep links, deployments, rollback, and mixed-version behavior.
- The supported terminal command is
mitsuro. The deprecatedkrustycommand remains a tested compatibility alias for the announced transition window; removing it requires the migration cutover criteria to pass. - If product language and an internal identifier differ, prefer clear translation at the UI/API boundary over leaking the internal name into the interface.
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 · 338 lines · 6,346 tokens per session scan A dea37f47e0d8
Mitsuro AGENTS.md is an instructions file published in the GitHub repository honeycomb-Technologies/Mitsuro (5 stars, last pushed 4d ago), licensed MIT. It adds 6,346 tokens to every session, about $0.0317 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
uniwind AGENTS.md
AGENTS.md instructions for uni-stack/uniwind, covering agent instructions and checks.
expo-targets AGENTS.md
AGENTS.md instructions for csark0812/expo-targets, covering agents, product posture, sealed zones, docs ssot and safe commands.
kiro-discord-bot AGENTS.md
AGENTS.md instructions for nczz/kiro-discord-bot, covering agents.md — kiro-discord-bot agent contract, source-of-truth map, non-negotiable principles, architecture boundaries and discord output and i18n.
zeno-mobile-runner AGENTS.md
Instructions for johnmikel/zeno-mobile-runner, covering notes for coding agents, public metadata policy — read before committing, the rule that catches agents out, scanned surfaces and enable the pre-push hook — once per clone.
ZenCODE AGENTS.md
AGENTS.md instructions for gerardogrisolini/ZenCODE, covering agents.md, project, architecture map, build and validation and linux validation.
opencode-mobile AGENTS.md
Instructions for Shahfarzane/opencode-mobile, covering openchamber - ai agent & contributor reference, tech stack, structure, where to look and architecture.