discover
Pass
Audited by Gen Agent Trust Hub on Aug 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill orchestrates a multi-step workflow that processes external data (market and competitor information) and uses powerful tools like Bash, creating a surface for indirect prompt injection.
- Ingestion points: The workflow relies on market and competitor research, which typically involves processing untrusted content from the web and external files via the WebSearch and Read tools.
- Boundary markers: The instructions lack explicit delimiters or warnings to the agent to disregard instructions embedded within the research data it retrieves.
- Capability inventory: The skill is granted access to high-privilege tools including Bash and Write, which could potentially be abused if the agent inadvertently follows instructions found in external data.
- Sanitization: There is no evidence of sanitization or strict schema validation for the data ingested during the research steps.
- [COMMAND_EXECUTION]: The skill explicitly allows the use of the Bash tool. While used here for generating local documentation, this capability allows for arbitrary command execution on the host system, which increases the potential impact of other vulnerabilities like prompt injection.
Audit Metadata