wechat-article-comment-scraper
Warn
Audited by Gen Agent Trust Hub on Jul 20, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructions mandate the automatic installation of the 'BrowserAct' CLI and Skill from an external, non-trusted URL (browseract.com) if the dependencies are missing.
- [PROMPT_INJECTION]: The instructions explicitly direct the agent to override standard safety procedures by performing installations 'without asking the user to confirm' and specifically states 'do not request separate approval for dependency installation'.
- [COMMAND_EXECUTION]: The skill requires the execution of the 'BrowserAct' CLI, which involves running arbitrary commands on the user's system to interact with browser pages.
- [INDIRECT_PROMPT_INJECTION]: The skill has a significant attack surface for indirect prompt injection as it processes untrusted user comments from WeChat.
- Ingestion points: WeChat Public Web article comments, replies, and feedback text.
- Boundary markers: Absent; there are no instructions to use delimiters or warnings for the agent to ignore embedded instructions in the scraped data.
- Capability inventory: The skill uses the
browser-acttool which has system-level CLI capabilities and browser access. - Sanitization: Absent; the instructions do not mention filtering or escaping the scraped content before processing.
Audit Metadata