dark-funnel

Pass

Audited by Gen Agent Trust Hub on Feb 27, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests and analyzes data from external, untrusted sources which could contain malicious instructions.
  • Ingestion points: Process step 2 identifies that the skill collects data from Reddit discussions, community forums, and customer survey responses.
  • Boundary markers: There are no delimiters or boundary markers specified in the process to separate the external data from the agent's internal instructions.
  • Capability inventory: The skill accesses the local filesystem to read brand profiles, guidelines, and SOPs from the ~/.claude-marketing/ directory as described in Process step 1.
  • Sanitization: The skill does not describe any methods for sanitizing, filtering, or validating the ingested content before it is processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 27, 2026, 01:02 PM