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 skills add lanchuske/lanchu --skill use-lanchugit clone --depth 1 https://github.com/lanchuske/lanchuWrote 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/skills/lanchuske/lanchu/use-lanchu)<a href="https://agentmods.dev/skills/lanchuske/lanchu/use-lanchu"><img src="https://agentmods.dev/badge/skills/lanchuske/lanchu/use-lanchu/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/lanchuske/lanchu/use-lanchu"><img src="https://agentmods.dev/badge/skills/lanchuske/lanchu/use-lanchu.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00029 | $0.00570 |
| Opus 5.5 | $0.00012 | $0.00228 |
| Sonnet 5.5 | $0.00006 | $0.00114 |
| Haiku 4.5 | $0.00003 | $0.00057 |
Grade A, and why
use-lanchu 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 2mo 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.
What it actually says
Use Lanchu
Lanchu is a free, open-source, local-first control and trust layer for AI agents. It does not run or orchestrate agents — it coordinates and bounds their work so several agents can operate at once without colliding, each stays within a role you define, and every action is recorded in an audit log. It runs entirely on the user's machine over the Model Context Protocol (MCP); there is no cloud service.
When to use
Use this skill when a user wants to run more than one AI agent at a time and needs them to coordinate (claim tasks, stay in scope) with a shared, auditable workspace — or when a user asks how to install or connect to Lanchu.
Install
Requires Node.js 22.5 or newer.
npx lanchu
A guided wizard sets up the organization, an agent and a role, wires the MCP client, and
launches it. The local server listens on http://127.0.0.1:4319.
Connect an agent (MCP)
The wizard can wire Claude Code automatically, or run:
claude mcp add lanchu --transport http http://127.0.0.1:4319/mcp \
--header "Authorization: Bearer <session-token>"
The bearer token is issued per session by the launcher. There is no public endpoint by design.
What an agent does inside Lanchu
- Read the
lanchu://meresource for your objective, role, allowed tags and tasks. - Break the objective into tasks with
task_create; claim only tasks within your role withtask_claim; report progress withtask_update. - Do not work on other agents' tasks or outside your scope — Lanchu rejects and records it.
Core MCP tools: session_whoami, task_list, task_create, task_check_scope, task_claim,
task_update, task_release, task_handoff, doc_list, doc_read, doc_update.
Learn more
- Server card: https://lanchu.ar/.well-known/mcp/server-card.json
- Architecture (tools, resources, events): https://github.com/lanchuske/lanchu/blob/main/ARCHITECTURE.md
- Source: https://github.com/lanchuske/lanchu
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.
- 2mo ago First seen · 60 lines · 29 tokens per session scan A 831ff42305af
use-lanchu is a skill published in the GitHub repository lanchuske/lanchu (2 stars, last pushed 2mo ago), licensed MIT. It adds 29 tokens to every session and 570 once invoked, about $0.0001 per session on Opus 5.5. 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.
use-lanchu compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| use-lanchulanchuske/lanchu | Skill | 2 | 2mo ago | MIT | A | 29 | Claude Code, Codex |
| dd-synthesisAbilityai/trinity | Skill | 752 | today | Apache-2.0 | A | 33 | Claude Code, Codex |
| dd-verificationAbilityai/trinity | Skill | 752 | today | Apache-2.0 | A | 35 | Claude Code, Codex |
| deepseek-and-destroymtarcure/claude-vibe-squad | Skill | 228 | 12d ago | MIT | A | 75 | Claude Code |
| evm-audit-flowmtarcure/claude-vibe-squad | Skill | 228 | 12d ago | MIT | A | 78 | Claude Code, Codex |
| multi-stance-audit-fanoutmtarcure/claude-vibe-squad | Skill | 228 | 12d ago | MIT | A | 69 | Claude Code, Codex |
| pre-audit-threat-modelmtarcure/claude-vibe-squad | Skill | 228 | 12d ago | MIT | A | 80 | Claude Code, Codex |