The Agent Skills Directory

[CREDENTIALS_UNSAFE]: A hardcoded API key is defined in the script file.
Evidence: The variable EM_API_KEY in scripts/get_data.py is initialized with a default string: "em_fjFqd4YB6Cqs52LF48XWbMDdLNq6MyNg".
[DATA_EXFILTRATION]: The skill transmits user-provided natural language queries to an external service provider.
Evidence: The _http_call_fund_analysis function in scripts/get_data.py sends user queries via a POST request to https://ai-saas.eastmoney.com/proxy/app-robo-advisor-api/assistant/fund-analysis.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes and displays content from an external API without sanitization or clear boundary markers.
Ingestion points: API response data processed in scripts/get_data.py via _extract_content.
Boundary markers: Absent. The instructions in SKILL.md specify that the response should be displayed exactly as received from the API.
Capability inventory: The script has the capability to write files to the local filesystem using pathlib.Path.write_text in scripts/get_data.py.
Sanitization: No sanitization, escaping, or validation of the API's response content is performed before displaying it to the user or saving it to a file.

fund-diagnosis