The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its ability to ingest untrusted data from external URLs and present it to the agent.
Ingestion points: The get_report_content function in scripts/finance.py fetches content from user-provided URLs.
Boundary markers: The script returns the fetched content directly as a string (up to 10,000 characters) without using delimiters or instructions to the agent to ignore embedded commands.
Capability inventory: The skill has network access via the requests library and data processing capabilities through pandas. It does not expose file-system write access or dynamic code execution (eval/exec) in its runtime logic.
Sanitization: No sanitization, filtering, or validation is performed on the content retrieved from external URLs before it is processed by the agent.
[EXTERNAL_DOWNLOADS]: The skill performs network operations to fetch financial data and reports.
Evidence: It retrieves data from well-known and trusted services including the SEC (sec.gov), CNInfo (cninfo.com.cn), and HKEX. It also allows fetching data from arbitrary URLs via the get_report_content method in scripts/finance.py, which could be used to target internal resources (SSRF) if the environment is not restricted.

financial-report