The Agent Skills Directory

[External Downloads] (MEDIUM): The skill executes pip install tushare pandas without version pinning or hash verification, allowing for the potential installation of untrusted or malicious versions of these libraries.
[Command Execution] (MEDIUM): The skill relies on the Bash(python:*) tool to execute Python code, granting the agent a high-privilege environment to process data.
[Prompt Injection] (HIGH): The skill exhibits a significant Indirect Prompt Injection vulnerability (Category 8) due to its interaction with external data. Evidence Chain: 1. Ingestion points: The skill reads from the Tushare Pro API (over 220 interfaces). 2. Boundary markers: Absent; there are no delimiters or warnings to ignore instructions inside the API response. 3. Capability inventory: Full Python execution via Bash(python:*). 4. Sanitization: Absent; the skill does not validate or sanitize the external data before processing it for the user.

tushare-finance