team-paid-media

Pass

Audited by Gen Agent Trust Hub on Mar 29, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the user to run the environment with the --allow-all flag (copilot --allow-all). This grants the agent full autonomy to execute commands without manual confirmation, which significantly increases the risk if the agent processes malicious input.
  • [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it processes external account data without safety boundaries.
  • Ingestion points: External CSV exports from Google Ads and Meta Ads specified in the prerequisites section.
  • Boundary markers: Absent; there are no instructions or delimiters used to prevent the agent from following commands potentially embedded in the CSV data.
  • Capability inventory: The skill manages a fleet of agents that read and write files to 'docs/paid-media/' and generate synthesized reports.
  • Sanitization: Absent; no validation or sanitization of the input data is mentioned before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 29, 2026, 02:51 PM
Security Audit — agent-trust-hub — team-paid-media