gtm-strategist
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill implements a stateful customization workflow that ingests user input and persists it within the skill's primary instruction file.
- Ingestion points: The agent is instructed to perform a 12-question 'Customization Interview' with the user, collecting data on business basics, tools, budget, and goals.
- Boundary markers: No delimiters or instructions are provided to the agent to treat the saved data as inert or to ignore potential instructions embedded within the user's responses.
- Capability inventory: The skill instructions explicitly authorize the agent to 'Save all changes to this file', which utilizes the agent's file-system write capabilities to modify its own source code (SKILL.md).
- Sanitization: There is no mention of sanitizing or validating the user's input before it is written into the SKILL.md file, potentially allowing a user to inject persistent instructions that would be executed in future sessions.
Audit Metadata