kai-monthly-audit
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The skill executes multiple internal Python scripts (e.g.,
scripts.audit.collect,scripts/ads/pull_all.py,scripts/quality_gates/audit_provenance_lint.py) to perform data collection, linting, and audit tasks within the agent's environment. - [INDIRECT_PROMPT_INJECTION]: The skill processes content from external websites and connected marketing APIs, which represents an attack surface for indirect prompt injection if the retrieved data contains malicious instructions.
- Ingestion points:
MARKETING.md, external website content via--url, and marketing platform data (GA4, GSC, CRM). - Boundary markers: The skill uses an automated provenance linting step (
audit_provenance_lint.py) to verify data sources, though explicit textual delimiters are not mentioned in the prompt instructions. - Capability inventory: The skill performs file writes to the local workspace and executes various Python-based auditing scripts.
- Sanitization: Content sanitization is not explicitly documented in the skill's instructions.
- [EXTERNAL_DOWNLOADS]: Fetches marketing and SEO data from various well-known external platforms and third-party providers (Google, DataForSEO) to populate the audit reports.
Audit Metadata