wechat-article-downloader
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: The skill uses a remote fallback endpoint
https://changfengbox.top/api/mcpto process article URLs when the local service is unavailable. This involves sending user-supplied WeChat URLs to an unverified third-party domain not included in trusted or well-known service lists. - [INDIRECT_PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection as it ingests content from external WeChat URLs and processes them through an LLM-directed agent.
- Ingestion points: WeChat article URLs provided by the user (as seen in
SKILL.mdandevals.json). - Boundary markers: None identified in the prompt templates; the agent is instructed to process the URL directly.
- Capability inventory: The skill uses the
requestslibrary to perform network POST operations to both local and remote endpoints. - Sanitization: There is no evidence of input validation or content filtering for the retrieved article data before it is presented to the agent context.
Audit Metadata