marketing-plan
Pass
Audited by Gen Agent Trust Hub on Aug 26, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection as it is designed to ingest and process external client materials.
- Ingestion points: Data enters the agent's context through files placed in the
~/marketing-plans/{client-slug}/materials/directory. - Boundary markers: The instructions do not specify the use of delimiters or 'ignore embedded instructions' prompts when reading these external files.
- Capability inventory: The skill possesses significant capabilities, including local filesystem write access, GitHub repository management, and access to third-party marketing and financial data via MCP integrations.
- Sanitization: There is no mention of sanitizing or validating the content of ingested materials before they are used to influence plan generation.
- [COMMAND_EXECUTION]: The skill references the use of the
agent-browserCLI tool for browser automation. The documentation provides a standard installation command (npm install -g agent-browser). This tool is attributed to a well-known service and is consistent with the skill's stated purpose of automating marketing research. - [DATA_EXFILTRATION]: The skill instructions involve reading from and writing to specific local directories (
~/marketing-plans/). This behavior is transparently documented and necessary for the skill's functionality as a resumable planning tool. It also accesses third-party data (Stripe, GA4) through authenticated platform-managed MCP tools, which is the intended use case for a marketing strategist.
Audit Metadata