fragments-policy

Warn

Audited by Gen Agent Trust Hub on Mar 5, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONREMOTE_CODE_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes npx fragments govern check --cloud to validate policies and interface with vendor services.\n- [REMOTE_CODE_EXECUTION]: The skill translates natural language descriptions into TypeScript code files within the .fragments/policies/ directory, which are subsequently executed using a CLI tool. This dynamic generation and execution of code from untrusted input is a significant behavior.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection where user instructions could maliciously influence the generated TypeScript logic.\n
  • Ingestion points: Design policy descriptions provided in create and edit modes.\n
  • Boundary markers: No delimiters or protective instructions are used to isolate the user input during code generation.\n
  • Capability inventory: Subprocess execution via npx, and file write/modification access to .fragments/ and fragments.config.ts.\n
  • Sanitization: No evidence of validation or sanitization of the natural language input before it is used to generate code.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Mar 5, 2026, 05:56 PM