blogwatcher

Warn

Audited by Gen Agent Trust Hub on Mar 12, 2026

Risk Level: MEDIUMEXTERNAL_DOWNLOADSPROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill installs the blogwatcher CLI tool directly from a third-party GitHub repository (github.com/Hyaxia/blogwatcher). This introduces a dependency on external code that is not from a trusted vendor or the skill's author.
  • [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface. 1. Ingestion point: External RSS/Atom feeds via the blogwatcher scan command (SKILL.md). 2. Boundary markers: Absent. 3. Capability inventory: Execution of CLI commands that return feed content to the agent context (SKILL.md). 4. Sanitization: Absent.
  • [DATA_EXFILTRATION]: The skill performs outbound network requests to fetch data from RSS/Atom feeds. Although intended for its functionality, it involves making connections to arbitrary external domains provided by users or feed sources.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 12, 2026, 02:37 AM