facebook-automation

Pass

Audited by Gen Agent Trust Hub on Sep 23, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external content from Facebook, such as user comments and posts. This creates a surface for potential indirect prompt injection attacks if the data being read contains malicious instructions intended to influence the agent's behavior.
  • Ingestion points: Facebook post content and user comments are read via the 'facebook' toolkit.
  • Capability inventory: The skill can write back to the environment by creating posts, replying to comments, and managing ad accounts.
  • Boundary markers: The provided instructions do not specify explicit delimiters for external data.
  • Sanitization: No specific sanitization logic is described in the prompt-level instructions.
  • [EXTERNAL_DOWNLOADS]: The skill requires the connection of the Rube MCP server located at 'https://rube.app/mcp'. This is a standard configuration step for using the Composio ecosystem for tool integration.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 23, 2026, 01:49 PM
Security Audit — agent-trust-hub — facebook-automation