Skill flagged — suspicious patterns detected

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

Find Arbitrage Opps

v1.0.0

Find arbitrage opportunities across exchanges by comparing prices for fungible token pairs like BTC/WBTC and USDT/USDC.

0· 397·1 current·1 all-time
byMichael Feng@fengtality
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Suspicious
View report →
OpenClawOpenClaw
Suspicious
medium confidence
Purpose & Capability
The code and SKILL.md match the stated purpose: they query a Hummingbot API for connectors, trading pairs, and prices to surface cross-exchange spreads. However the registry metadata declared no required environment variables/credentials while the SKILL.md and script rely on HUMMINGBOT_API_URL, API_USER, API_PASS and expect exchange API keys configured in Hummingbot. That metadata omission is an inconsistency.
!
Instruction Scope
SKILL.md instructs you to run a remote script via `bash <(curl -s https://raw.githubusercontent.com/...)` (executes code fetched from the network). The included Python script also reads local .env files (./hummingbot-api/.env, ~/.hummingbot/.env, .env) to load credentials. While reading those .env files is consistent with needing Hummingbot API credentials, instructing remote script execution and reading arbitrary .env locations expand the scope and risk.
!
Install Mechanism
There is no declared install spec (instruction-only), but SKILL.md explicitly tells users to execute a fetched script from raw.githubusercontent.com. Executing code directly from the network is higher risk and should be audited before running. The skill itself (the Python file) has no build/install steps.
!
Credentials
The registry claims no required env vars, but both SKILL.md and scripts use HUMMINGBOT_API_URL, API_USER, API_PASS and expect exchange connector API keys configured in Hummingbot. The script will load local .env files (which can contain unrelated secrets). Defaults to admin/admin are insecure. Required secrets are proportionate to the task, but the metadata mismatch and automatic .env loading are concerning.
Persistence & Privilege
The skill does not request persistent presence (always:false) and does not modify other skills or system-wide config. It only runs as a script/command and uses the Hummingbot API. No elevated platform privileges requested.
What to consider before installing
This skill appears to implement what it claims (queries a local Hummingbot API for prices and finds spreads), but there are red flags you should address before installing or running it: - Do NOT blindly run the suggested `bash <(curl -s https://raw.githubusercontent.com/...)` command. That fetches and executes a remote script; inspect its contents first (view the URL in a browser or curl it to a file and read it). - The registry metadata says no env vars, but the script expects HUMMINGBOT_API_URL, API_USER, API_PASS and that exchange connectors are configured with API keys. Provide only least-privileged credentials and avoid using default admin/admin. - The Python script reads .env files from common Hummingbot locations. Those files can contain secrets; verify which .env will be loaded and ensure it doesn't expose unrelated credentials. - Network scope: the script only contacts the configured HUMMINGBOT_API_URL endpoints (connectors, trading-rules, market-data/prices) and uses HTTP Basic auth. Ensure that HUMMINGBOT_API_URL points to a trusted/local service (e.g., localhost) and that the API user has limited permissions. - If you’re unsure, run the script in an isolated environment/container with no sensitive .env mounted, inspect the script code (already included), and manually perform any prerequisite checks rather than executing remote installers. If you want, I can: (a) fetch and show the contents of the referenced prerequisite script URL for review, (b) summarize the full Python file (it was truncated in input) to confirm there are no hidden exfiltration routines, or (c) suggest a safe run checklist/command set to test the skill in isolation.

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

latestvk97bc6gdyjp04pmsqs64wkp4y981y98g

License

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

Comments