jwt-encode

Pass

Audited by Gen Agent Trust Hub on Apr 7, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill constructs and executes shell commands using node -e, python3 -c, and openssl to sign JWT payloads.
  • [EXTERNAL_DOWNLOADS]: The skill downloads and installs the 'jose' package from the official NPM registry if it is not detected in the environment.
  • [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by interpolating user-provided claims directly into shell command templates without validation.
  • Ingestion points: Claims and payload data provided by the user in SKILL.md.
  • Boundary markers: Absent; data is placed directly into script strings.
  • Capability inventory: Shell command execution (Node.js, Python, OpenSSL) as defined in SKILL.md.
  • Sanitization: No escaping or validation is performed on the input claims before they are passed to the shell.
  • [SAFE]: The skill uses environment variables (JWT_SECRET) to handle sensitive keys, which prevents exposure in system process lists or shell history.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 7, 2026, 04:01 PM