linkedin-comment-drafter

Pass

Audited by Gen Agent Trust Hub on Sep 6, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external LinkedIn posts and comments.
  • Ingestion points: The skill fetches data via lib.ApifyClient.fetch_post and fetch_post_comments as described in SKILL.md.
  • Boundary markers: SKILL.md includes a specific 'Untrusted content' section with instructions to treat fetched data as data and not instructions, which acts as a defense against injection.
  • Capability inventory: The skill can execute lib.publish and lib.repost (commenting and resharing) based on the fetched data, as specified in SKILL.md.
  • Sanitization: The skill uses a humanizer pass and predefined templates from references/comment-templates.md to constrain and format the output generated from external content.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 6, 2026, 12:19 PM
Security Audit — agent-trust-hub — linkedin-comment-drafter