audience-intelligence

Warn

Audited by Gen Agent Trust Hub on Feb 27, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructions in SKILL.md specify executing a local Python script via the command python campaign-tracker.py --brand {slug} --action list-campaigns. The {slug} variable is interpolated directly into the shell command without evidence of validation or sanitization, which could lead to command injection if an attacker can manipulate the brand name or slug.\n- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection due to its workflow of ingesting and processing untrusted data from external sources.\n
  • Ingestion points: The skill reads brand profiles and guidelines from ~/.claude-marketing/ and is designed to process external research data such as CRM exports, survey results, and interview transcripts.\n
  • Boundary markers: There are no markers or delimiters defined to help the agent distinguish between data and instructions within the ingested files.\n
  • Capability inventory: The agent possesses capabilities to read/write files and execute shell commands (via Python).\n
  • Sanitization: There is no mention of sanitization, escaping, or validation of the content within the data files or external research inputs.\n- [EXTERNAL_DOWNLOADS]: The skill references multiple external files and a script (campaign-tracker.py) that are not included in the provided package. This creates a dependency on unverified external code and configurations, the security of which cannot be evaluated from the skill's own file set.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 27, 2026, 01:02 PM