setup

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill asks the user for API keys/secrets and instructs the agent to store them by writing them directly into a .env file (and to not ask the user to edit it manually), which requires the LLM to receive and output secret values verbatim.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The setup script explicitly downloads and builds remote code during runtime (wget http://prdownloads.sourceforge.net/ta-lib/ta-lib-0.4.0-src.tar.gz and an optional wheel from https://github.com/cgohlke/talib-build/releases), which fetches and executes external code as a required installation step for TA-Lib.

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

• Direct money access detected (high risk: 1.00). The skill explicitly installs and references crypto/exchange tooling and API keys: it adds ccxt to the pip install list and asks the user to provide/store Binance API key & secret and an OpenAlgo API key in the .env. ccxt and exchange API keys are specific crypto/trading integrations that enable authenticated trading/market actions (direct financial execution), so this skill enables direct financial execution capability.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 0.90). This skill explicitly instructs the agent/user to run OS-level installation commands that require sudo (apt/yum, make install into /usr, brew install), which modify system files and request elevated privileges.