ad-creative
Pass
Audited by Gen Agent Trust Hub on Feb 25, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill implements an 'Iterate from Performance Data' mode that processes external data from CSVs, text pastes, or API outputs. This creates a surface for indirect prompt injection where malicious instructions could be embedded in the performance data to manipulate the agent's creative generation or influence tool interactions.
- Ingestion points: 'Mode 2: Iterate from Performance Data' in SKILL.md.
- Boundary markers: The skill does not define explicit delimiters or instructions for the agent to ignore instructions embedded within the provided data.
- Capability inventory: The skill utilizes CLI tools for Google Ads, Meta Ads, LinkedIn Ads, and TikTok Ads, and can generate structured CSV files.
- Sanitization: No sanitization or validation of the input performance data is specified.
- [COMMAND_EXECUTION]: The reference documentation (references/generative-tools.md) includes specific shell commands for tool setup and API interaction, such as 'git clone', 'make', and 'curl'. While provided as guidance for the user, these instructions could be executed by an agent in environments where these commands are available.
Audit Metadata