lov-align-wechat-uiux
Pass
Audited by Gen Agent Trust Hub on Oct 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data sources, creating a potential vector for indirect prompt injection.\n
- Ingestion points: As described in
SKILL.mdandskills/wechat-semantics/SKILL.md, the skill processes raw WeChat message data, including complex XML structures, JSON payloads, and database records.\n - Boundary markers: The instructions do not establish explicit boundary markers or "ignore" directives to isolate the external raw data from the agent's primary command context, which could lead to the agent inadvertently executing instructions hidden within the data.\n
- Capability inventory: The skill possesses capabilities to inspect the repository, modify source code for UI repairs, and execute local validation scripts (
scripts/validate_parity_case.py).\n - Sanitization: There is no evidence of sanitization or filtering logic applied to the external raw records before they are interpreted for semantic modeling.
Audit Metadata