marketing-plan
Pass
Audited by Gen Agent Trust Hub on Jul 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a legitimate marketing strategy workflow. It manages its own state using a local directory structure (
~/marketing-plans/) and uses platform-standard Model Context Protocol (MCP) integrations for tools like Stripe, GitHub, and Ahrefs. No malicious code execution or credential theft was found. - [DATA_EXPOSURE_&_EXFILTRATION]: While the skill interacts with sensitive business data via MCPs (Stripe, GA4, Ahrefs), it does so as a core functional requirement. It does not attempt to access unauthorized system files (e.g., SSH keys or AWS credentials) or exfiltrate data to unverified third-party domains. The 'publish to GitHub' feature is user-initiated and follows standard authorization patterns.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses a vulnerability surface for indirect prompt injection as it ingests untrusted client materials.
- Ingestion points: Files located in the
~/marketing-plans/{client-slug}/materials/directory are read during the INIT phase. - Boundary markers: Absent; the instructions do not explicitly mandate delimiters or 'ignore embedded instructions' warnings when reading these files.
- Capability inventory: The agent has permissions to write files to the local disk and interact with several external APIs via MCP (Stripe, GitHub, Notion, etc.).
- Sanitization: Absent; no validation or escaping of external content before interpolation is described.
Audit Metadata