reg-feed-watcher
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill is configured to fetch regulatory data from a wide variety of official and reputable sources. These include Chinese government portals (e.g., gov.cn, cac.gov.cn, miit.gov.cn), US federal agencies (e.g., federalregister.gov, sec.gov, ftc.gov), and European Union institutions (e.g., ec.europa.eu, ico.org.uk). These are established, well-known organizations.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external regulatory feeds and websites, which represents an indirect prompt injection surface. However, this is inherent to the skill's primary function of monitoring the web for legal updates.
- Ingestion points: Step 1 pulls data from numerous external URLs and RSS/JSON feeds defined in SKILL.md and references/source-catalog.md.
- Boundary markers: The instructions do not define specific delimiters or instructions to ignore embedded directives in the fetched content.
- Capability inventory: The skill performs network read operations, reads local configuration files from ~/.claude/plugins/config/, and writes summary reports to the file system.
- Sanitization: The skill does not explicitly sanitize the fetched text, but it implements a tagging system ([联网检索 — 需复核], [模型知识 — 需验证]) to alert the user to the source and reliability of the data.
Audit Metadata