linkfox-zhihuiya-fulltext-image
Fail
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: HIGHPERSISTENCEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [PERSISTENCE]: The
references/onboarding.mdfile provides instructions to append environment variable exports to shell profile files such as~/.bashrcand~/.zshrcto maintain access across sessions. - [EXTERNAL_DOWNLOADS]: The
scripts/onboarding.pyfile requires non-standard Python libraries includingrequests,qrcode, andpillowfor handling web requests and generating payment QR codes. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from patent identifiers and user-provided authentication data which could lead to prompt injection vulnerabilities.
- Ingestion points: Patent IDs and publication numbers in
scripts/zhihuiya_fulltext_image.py; phone numbers and login codes inscripts/onboarding.py. - Boundary markers: The skill lacks explicit delimiters or instructions to ignore embedded commands in the data it processes.
- Capability inventory: The skill performs network requests to remote APIs and writes metadata and results to the local file system.
- Sanitization: Input parameters are not strictly validated or sanitized before being transmitted to external APIs or used in file path generation.
- [DYNAMIC_EXECUTION]: The
scripts/zhihuiya_fulltext_image.pyscript manipulatessys.pathto dynamically load modules from computed relative paths (../../_shared).
Recommendations
- AI detected serious security threats
Audit Metadata