facebook-marketing

Pass

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests untrusted user data to generate marketing content and code scripts.
  • Ingestion points: User prompts containing business names, product descriptions, and campaign goals as seen in Example 1 and Example 2 (SKILL.md).
  • Boundary markers: The skill lacks explicit instructions or delimiters to ignore embedded commands within the user-provided data.
  • Capability inventory: The skill generates executable TypeScript code that performs network operations (API publishing) and file management (Page/Group management).
  • Sanitization: No explicit escaping or validation of user-provided strings is recommended or implemented before interpolation into generated scripts or posts.
  • [EXTERNAL_DOWNLOADS]: The provided code templates interact with Facebook's official infrastructure.
  • Code snippets demonstrate network requests to graph.facebook.com and www.facebook.com for authentication, publishing content, and retrieving analytics.
  • [CREDENTIALS_UNSAFE]: The skill follows security best practices for credential management by using environment variables.
  • Code templates utilize process.env.FB_APP_ID, process.env.FB_APP_SECRET, and process.env.FB_PAGE_TOKEN rather than hardcoding sensitive tokens.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 13, 2026, 09:15 PM