build-my-skill
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user data, including conversation history, process drafts, and templates, to generate instructions for a new AI skill. This creates a surface where a malicious user could provide input designed to influence the behavior of the resulting skill.
- Ingestion points: Processes conversation history and user-supplied drafts, templates, and examples as defined in
SKILL.md. - Boundary markers: Absent. The instructions do not specify the use of delimiters or specific markers to isolate user-supplied content from the instruction generation logic.
- Capability inventory: The skill possesses file-write capabilities for installation and mentions script execution for testing purposes in
references/build-and-install.md. - Sanitization: Absent. The skill is designed to faithfully capture and encode the user's process, relying on the agent's general safety filters for content generation.
- [COMMAND_EXECUTION]: The instructions in
references/build-and-install.mddirect the agent to use native system tools or official installation commands to manage the local skill directory and install the generated files. This capability is necessary for the skill's primary function of installing generated workflows.
Audit Metadata