Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Openclaw Quant Skill
v0.1.0Professional quantitative trading system for cryptocurrency - backtesting, paper trading, live trading, and strategy optimization
⭐ 0· 309·0 current·0 all-time
byJustin Liu@zhenstaff
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
The SKILL.md describes a crypto quant trading system (backtest/paper/live trading) which reasonably needs exchange API keys and access to external repos, but one of the included files (SKILL-EN.md) is for a completely different 'video-generator' skill (front matter name: video-generator). Including unrelated skill docs in the same package is an inconsistency that could indicate sloppy packaging or intentional misdirection.
Instruction Scope
Instructions tell the agent to clone external GitHub repos and run pip/npm installs and CLI commands; they reference environment variables (BINANCE_API_KEY, BINANCE_API_SECRET and, in the video doc, OPENAI_API_KEY) and include examples for running live trades. The skill metadata declares no required env vars, yet the runtime instructions assume API credentials for live trading — the agent could be directed to perform actions (including real trades) that require secrets not declared in the manifest.
Install Mechanism
This is instruction-only (no install spec), so nothing is written by the skill itself. However, the instructions direct the user/agent to git clone repositories and run pip/npm installs from external GitHub repos, which will execute/introduce third-party code. That behavior is expected for such tools but increases risk and requires verifying the upstream repository before installation.
Credentials
The manifest lists no required environment variables/primary credential, yet SKILL.md references BINANCE_API_KEY/BINANCE_API_SECRET for live trading and SKILL-EN references OPENAI_API_KEY. A trading skill that can place real orders should explicitly declare required credentials and document least-privilege usage (testnet keys, withdraw-disabled keys). The absence of declared credentials is disproportionate and reduces transparency.
Persistence & Privilege
always:false (no forced inclusion) which is normal. The skill allows autonomous invocation (platform default). Combined with the ability to trigger live trading, autonomous invocation increases blast radius — verify confirmation behavior and that the agent will not place live orders without explicit user consent.
What to consider before installing
This package is internally inconsistent and needs verification before use. Recommended actions: 1) Do not run commands that clone or pip/npm install the repositories until you inspect the upstream GitHub projects (https://github.com/ZhenRobotics/openclaw-quant and the video repo referenced) and review their code. 2) Ask the publisher to fix the packaging (remove unrelated SKILL-EN.md or explain why it’s included). 3) If you plan to use live trading, only use API keys with minimal permissions (no withdraw permission) and prefer testnet keys; require explicit confirmation before any live order. 4) Start with paper trading and audit logs/commands the agent would run. 5) If you’re not comfortable reviewing the repos, avoid installing and consider a vetted alternative. 6) Request an updated manifest that explicitly lists required environment variables and clearly documents the live-trading confirmation flow.Like a lobster shell, security has layers — review code before you run it.
latestvk9797jr86aj0mb7rc4h5jm9m9d82awxp
License
MIT-0
Free to use, modify, and redistribute. No attribution required.