Engagement Optimizer

Pass

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [NO_CODE] (SAFE): The skill contains only Markdown documentation and metadata. No executable scripts, binaries, or configuration files for package managers were found.
  • [REMOTE_CODE_EXECUTION] (SAFE): No patterns indicating the download or execution of remote scripts (e.g., curl | bash) were detected.
  • [DATA_EXFILTRATION] (SAFE): There are no hardcoded credentials, sensitive file path references, or network-enabled commands that could be used for data theft.
  • [Indirect Prompt Injection] (LOW): The skill defines a strategy for processing external social media data (comments and DMs) for automation purposes without defining security boundaries.
  • Ingestion points: External social media comments and Direct Messages (implied by 'comment strategies' and 'DM automation').
  • Boundary markers: None present; the instructions do not specify delimiters or warnings to ignore embedded instructions in external text.
  • Capability inventory: Implied capabilities include reading/writing social media comments and sending automated DMs.
  • Sanitization: No sanitization, escaping, or validation steps are provided for the content ingested from external users.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 17, 2026, 05:19 PM