skills/mr-kelly/skills/kelly-social/Gen Agent Trust Hub

kelly-social

Pass

Audited by Gen Agent Trust Hub on Aug 25, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted external content from social media platforms (X, Facebook, Instagram) which creates a surface for indirect prompt injection attacks.\n
  • Ingestion points: Ingests social media mentions and comments through scripts/ingest_snapshot.mjs populated via browser automation or analytics exports.\n
  • Boundary markers: The skill uses a social-qa quality gate in social-model.js to evaluate outgoing content, and SKILL.md defines boundaries for data collection.\n
  • Capability inventory: The agent can draft posts, replies, and scripts, and perform browser automation for data collection.\n
  • Sanitization: The skill normalizes metrics and validates outgoing drafts, but it does not explicitly sanitize incoming social media text against injection patterns before the agent processes it for drafting replies.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 25, 2026, 10:36 AM
Security Audit — agent-trust-hub — kelly-social