kai-retention
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local Python scripts (
scripts/quality_gates/banned_word_check.pyandscripts/quality_gates/four_us_score.py) to perform quality checks on generated content.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes various project files to automatically generate product context, creating a surface for malicious data to influence agent behavior.\n - Ingestion points: Files like
MARKETING.md,README.md,package.json, and general codebase files.\n - Boundary markers: No specific delimiters or safety instructions are provided to the agent to treat this data as untrusted.\n
- Capability inventory: The skill has the ability to read/write files and execute Python scripts via the command line.\n
- Sanitization: There is no mention of sanitizing or validating the content extracted from project files.\n- [DATA_EXFILTRATION]: The skill directs the agent to search for and read email, ad, and analytics configuration files, which are common locations for sensitive API keys, tokens, and credentials.
Audit Metadata