cli-creator
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- Indirect Prompt Injection (Category 8): The skill's core functionality involves ingesting and analyzing untrusted external data (such as API documentation, OpenAPI specifications, or web application curls) to generate executable code. This is an inherent attack surface where malicious content in the source material could attempt to inject instructions into the generated CLI or the agent's logic.
- Dynamic Code Generation and Execution: The skill facilitates the creation of new software in Rust, TypeScript, or Python, which is then compiled and installed onto the user's system (e.g., in
~/.local/bin). This involves local command execution and modification of the environment, which are necessary for the skill's utility but should be monitored for unexpected behavior. - External Dependency Ingestion: The instructions suggest the use of common package managers (Cargo, npm, pip) to install dependencies for the generated tools. While these are standard tools, they involve downloading external code at runtime.
- Credential Handling Best Practices: The skill proactively addresses data exposure risks by requiring the redaction of tokens and cookies from outputs and recommending the use of environment variables or local config files over command-line flags to prevent secret leakage in process listings or shell history.
Audit Metadata