Skill flagged — suspicious patterns detected

ClawHub Security flagged this skill as suspicious. Review the scan results before using.

Vantage — HL Autonomous Trading Agent

v1.0.0

Vantage — Autonomous trading agent for Hyperliquid perpetual futures. Signal-to-execution in one loop. Runs on your machine. No cloud infra. No ongoing cost...

0· 189·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Suspicious
medium confidence
!
Purpose & Capability
The code and SKILL.md describe a local autonomous trading agent that requires a Hyperliquid private key and wallet address — which is coherent for the stated purpose. However the registry metadata (required env vars / primary credential) incorrectly lists none. That mismatch is significant because it hides the fact that a high-value secret (private key) is required.
Instruction Scope
Runtime instructions and code stay within trading-related scope: market data fetches (Hyperliquid, Midgard/THORNode, CoinGecko), local paper-trading state, LLM decision layer (local Ollama preferred, OpenAI optional), and order signing/execution. The SKILL.md and README instruct storing the private key in .env and running setup-check.js — expected for a trading bot. There is no obvious instruction to read unrelated files, but you should inspect trader.js and setup-check.js (not fully shown) to confirm the private key is only used locally and not sent to third parties.
Install Mechanism
No remote install/download step is present in the skill bundle; dependencies are standard npm packages (axios, ethers, dotenv, etc.). The package files are included in the bundle rather than being downloaded from arbitrary URLs, so install risk is typical for local Node code.
!
Credentials
The skill requires HYPERLIQUID_PRIVATE_KEY and HYPERLIQUID_WALLET_ADDRESS (high-value credentials) per SKILL.md and README — appropriate for trading — but the registry metadata does not declare them. Optional envs include OPENAI_API_KEY and OLLAMA_URL which enable remote/local LLMs. The omission of required env declarations in metadata reduces transparency and increases risk that users install without realizing they'll provide a private key.
Persistence & Privilege
The skill is not marked always:true and does not request elevated system-wide privileges. It writes a local paper-trades.json under the skill's data directory and otherwise behaves as a normal local CLI application.
What to consider before installing
This package is internally consistent with being a local trading bot — it legitimately needs your Hyperliquid private key to sign orders — but the registry metadata fails to declare those required secrets. Before using with real funds: (1) Do NOT supply your main/private wallet key until you audit trader.js and setup-check.js to confirm the key is only used locally and never transmitted. Search the codebase for any network calls that include process.env.HYPERLIQUID_PRIVATE_KEY or other sensitive data. (2) Run entirely in --paper mode first and verify one complete cycle. (3) Prefer using a funded test account or a wallet with minimal funds. (4) Avoid setting OPENAI_API_KEY unless you understand remote LLM fallback behavior (it will call api.openai.com). (5) Ask the skill author or registry to correct the metadata to explicitly require HYPERLIQUID_PRIVATE_KEY and HYPERLIQUID_WALLET_ADDRESS so the requirement is visible at install time. If you cannot review trader.js or cannot confirm the key is never exfiltrated, do not run live trading with valuable keys.
src/hyperliquid.js:14
Environment variable access combined with network send.
src/index.js:82
Environment variable access combined with network send.
src/setup-check.js:105
Environment variable access combined with network send.
src/sizing.js:32
Environment variable access combined with network send.
src/thorchain.js:11
Environment variable access combined with network send.
!
src/hyperliquid.js:123
File read combined with network send (possible exfiltration).
Patterns worth reviewing
These patterns may indicate risky behavior. Check the VirusTotal and OpenClaw results above for context-aware analysis before installing.

Like a lobster shell, security has layers — review code before you run it.

latestvk975nks35h260b17n2bppsmrdh82xtd3

License

MIT-0
Free to use, modify, and redistribute. No attribution required.

Runtime requirements

📈 Clawdis
Binsnode

Comments