replitica-grow

Pass

Audited by Gen Agent Trust Hub on Jun 16, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill processes external content from social media posts, which presents an indirect prompt injection surface.
  • Ingestion points: Untrusted data from X/Twitter is processed via the list_opportunities and get_opportunity tools as defined in SKILL.md.
  • Boundary markers: The instructions do not define specific delimiters or "ignore instructions" markers for the external post content.
  • Capability inventory: The skill allows the agent to draft and schedule content and toggle account autopilot settings using the provided MCP tools.
  • Sanitization: No content sanitization or filtering is described for the retrieved post data.
  • Mitigation: The core logic requires explicit user confirmation and a "clear yes" before any reply is scheduled, ensuring human review of any content potentially influenced by external input.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 16, 2026, 09:56 AM
Security Audit — agent-trust-hub — replitica-grow