comment-funnel

Pass

Audited by Gen Agent Trust Hub on Jul 13, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill requires the agent to ingest and evaluate untrusted text from Instagram comments to trigger automated DMs, creating a surface for indirect prompt injection. \n
  • Ingestion points: External comment data is retrieved via the GET /uploadposts/comments API endpoint as described in SKILL.md. \n
  • Boundary markers: The skill lacks explicit instructions to use delimiters or other boundary markers to isolate untrusted user content from the agent's instructions. \n
  • Capability inventory: The agent has the capability to send private messages and initiate automated monitoring processes based on the content of the processed comments. \n
  • Sanitization: The instructions rely on the agent's semantic judgment to identify intent and do not specify strict input validation or sanitization protocols for the external data.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 13, 2026, 01:52 PM
Security Audit — agent-trust-hub — comment-funnel