linkedin-outreach

Pass

Audited by Gen Agent Trust Hub on Mar 14, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill implements an indirect prompt injection surface by interpolating untrusted data into its message generation logic.
  • Ingestion points: The skill retrieves external data from the Supabase people table, specifically utilizing the {comment_snippet} variable which contains raw text from LinkedIn comments.
  • Boundary markers: There are no boundary markers or specific instructions (e.g., XML tags or 'ignore' directives) surrounding the interpolated variables to differentiate data from instructions.
  • Capability inventory: Across its scripts, the skill has the capability to write to the local filesystem (skills/linkedin-outreach/output/), write to Google Sheets via the google-sheets-write tool, and perform authenticated network operations (PATCH/POST) to Supabase.
  • Sanitization: The skill does not perform any sanitization, escaping, or validation of the retrieved signal data before it is presented to the LLM for message construction.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 14, 2026, 06:02 PM