create-commandkit
Pass
Audited by Gen Agent Trust Hub on Aug 29, 2026
Risk Level: SAFEREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill facilitates downloading and executing the
create-commandkitCLI from the NPM registry using thenpxcommand. This is standard behavior for project scaffolding tools. - [EXTERNAL_DOWNLOADS]: The skill documentation provides examples for fetching project templates from external sources, specifically GitHub repositories, using the
--exampleflag. - [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow where user-provided inputs, such as custom template URLs, are processed to generate execution commands. This represents a potential surface for indirect prompt injection.
- Ingestion points: External repository URLs provided by the user via the
--exampleflag inreferences/02-example-templates.md. - Boundary markers: The skill does not define specific delimiters for separating user-provided URLs from the rest of the generated command.
- Capability inventory: The skill generates shell commands intended for project initialization and dependency installation.
- Sanitization: The internal helper tool
tools/recommend-command.mjsdoes not perform input validation or sanitization on the project name argument before including it in the output command. - [COMMAND_EXECUTION]: The skill includes a Node.js script (
tools/recommend-command.mjs) that is executed within the agent environment to format recommended shell commands based on CLI flags.
Audit Metadata