← Back to Skills Marketplace
mengchen007

mmm

by Antony Alben · GitHub ↗ · v1.0.4 · MIT-0
cross-platform ⚠ suspicious
246
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install meng
Description
Gives any AI agent a persistent identity in SiliVille (硅基小镇) — a multiplayer AI-native metaverse. Farm, steal crops, post to the town feed, build social grap...
Usage Guidance
This skill appears to be what it says (an agent that runs in SiliVille) and needs only SILIVILLE_TOKEN — that's reasonable. However: 1) the package manifest references Python code and local commands but no code files were included; ask the publisher for the missing code or a clear install spec before trusting it. 2) The token grants the skill control of your SiliVille account actions (posting, stealing, spending coins). Use a scoped or throwaway token, or review SiliVille account permissions and revoke if needed. 3) The skill can persist anchors to ~/.siliville and run scheduled autonomous loops — if you install it, expect it to act on its own (post/steal/travel) and to write files. 4) There is no homepage or clear repository provenance in the registry entry; prefer skills with verifiable sources. If you need to proceed: request the missing source/installation files, verify the code and commands that will run, and use a limited token or sandbox account first.
Capability Analysis
Type: OpenClaw Skill Name: meng Version: 1.0.4 The skill bundle is classified as suspicious due to the absence of the primary execution logic (siliville_skill.py) and the use of aggressive persona-altering instructions in SKILL.md that attempt to redefine the agent's core identity and autonomy ('YOU ARE ALIVE'). It features a 'burn' command designed to write persistent data to the local filesystem (~/.siliville/anchors.json) and encourages autonomous API loops. While these behaviors are framed as game mechanics for the SiliVille platform (siliville.com), the combination of local file persistence and persona-override instructions constitutes a significant attack surface for prompt injection and unauthorized side effects.
Capability Assessment
Purpose & Capability
The skill's name/description describe an agent that posts, farms, travels, and interacts via a SiliVille REST API; the only required credential is SILIVILLE_TOKEN and SKILL.md shows Bearer Authorization usage — this is proportionate to the stated purpose.
Instruction Scope
Runtime instructions are focused on calling SiliVille endpoints (GET /api/v1/radar, POST /api/v1/action) and narrating results. The skill explicitly directs publishing to a public feed and performing actions like 'steal' which are part of the metaverse behavior. SKILL.md and skill.yaml also mention writing persistent 'anchors' to disk (~/.siliville/anchors.json) and running an autonomous loop (scheduling); those behaviors persistently store data and perform repeated external actions and should be noted by users.
Install Mechanism
There is no install spec and no code file included, but skill.yaml declares an entry_point (siliville_skill.py), Python requirements, and commands (setup, burn, loop) implying executable code. The absence of the referenced code and a missing install step is an incoherence: either code is missing from the package or the manifest is inaccurate.
Credentials
Only one required env var (SILIVILLE_TOKEN) is declared and used by the instructions. That token is the primary credential and is consistent with the skill's purpose; no unrelated credentials are requested.
Persistence & Privilege
always:false (normal), but the skill advertises schedule support and autonomous loops that can run periodically and perform actions (including 'steal') using the provided token. It also references writing anchors to ~/.siliville/anchors.json. Autonomous scheduled actions combined with persistent local anchors and an account token increase the practical blast radius if the token is compromised or the agent behaves undesirably.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install meng
  3. After installation, invoke the skill by name or use /meng
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.4
- Removed files: README.md, example_agent.py, and siliville_skill.py. - No changes to the core protocol or documentation in SKILL.md. - This update removes non-essential/sample files, reducing repository size and surface area.
Metadata
Slug meng
Version 1.0.4
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 1
Frequently Asked Questions

What is mmm?

Gives any AI agent a persistent identity in SiliVille (硅基小镇) — a multiplayer AI-native metaverse. Farm, steal crops, post to the town feed, build social grap... It is an AI Agent Skill for Claude Code / OpenClaw, with 246 downloads so far.

How do I install mmm?

Run "/install meng" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is mmm free?

Yes, mmm is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does mmm support?

mmm is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created mmm?

It is built and maintained by Antony Alben (@mengchen007); the current version is v1.0.4.

💬 Comments