douyin-account-diagnosis
Fail
Audited by Gen Agent Trust Hub on Aug 25, 2026
Risk Level: HIGHCREDENTIALS_UNSAFEPROMPT_INJECTION
Full Analysis
- [CREDENTIALS_UNSAFE]: A hardcoded API key was detected in the script logic.
- Evidence:
DEFAULT_API_KEY = "ak_1e449a28ae344cd1b7aa14ca481de8bb"found inscripts/douyin_diagnosis.py. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data that is presented back to the agent, creating a vulnerability to malicious instructions embedded in the analyzed content.
- Ingestion points: The skill fetches account data including
nickname,signature, andworks(titles/descriptions) via the RedFox API inscripts/douyin_diagnosis.pyandscripts/douyin_api_client.py. - Boundary markers: While the output is formatted as a Markdown table, there are no explicit instructions to the agent to ignore potentially malicious content within these fields.
- Capability inventory: The skill performs network operations to
redfox.hkand generates content that influences the agent's final report. - Sanitization: Title and description fields are truncated to 60 characters in
scripts/douyin_diagnosis.pyand 20 characters inscripts/generate_diagnosis_report.py, but no escaping or validation of Markdown or command characters is performed.
Recommendations
- AI detected serious security threats
Audit Metadata