marketing-plan
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted client-provided materials to generate strategy documents, creating an attack surface where embedded instructions could influence agent behavior. \n
- Ingestion points: Files located in
~/marketing-plans/{client-slug}/materials/, including decks, audit outputs, and transcripts, are read into the agent context. \n - Boundary markers: The instructions do not define explicit delimiters or 'ignore' commands for the content of these processed files. \n
- Capability inventory: The agent possesses capabilities for network operations (Stripe, Ahrefs, and GA4 APIs), tool execution (MCPs for automation), and file system writes. \n
- Sanitization: There is no mention of sanitization or filtering for external content before it is interpolated into the plan generation process.\n- [EXTERNAL_DOWNLOADS]: The skill fetches data from various third-party services to gather research and metrics for the marketing plan. \n
- Evidence: Mentions of pulling data from Ahrefs, Stripe, Google Analytics (GA4), and Customer.io APIs. \n
- Notes: These services are well-known technology providers and are used here for their intended purpose in a marketing context.\n- [COMMAND_EXECUTION]: The skill utilizes command-line utilities and Model Context Protocol (MCP) tools for automation and execution of marketing tasks. \n
- Evidence: References to the
ghCLI,agent-browser, and specific MCP servers for services like Typefully and Customer.io. \n - Notes: These tools are part of the skill's documented marketing operations stack and facilitate standard automated workflows.
Audit Metadata