jwt-oauth-token-attacks

Fail

Audited by Gen Agent Trust Hub on Nov 9, 2026

Risk Level: CRITICALCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides numerous shell commands for decoding JWTs, generating attack payloads, and executing security tools such as jwt_tool.py, hashcat, and john. These commands are intended for security professional workflows but involve direct execution of shell utilities.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external authentication tokens (JWTs) provided by a user or target environment. This creates a vulnerability surface where a maliciously crafted token could attempt to influence the agent's behavior during analysis.
  • Ingestion points: Authentication tokens provided via placeholders like JWT_TOKEN or JWT_TOKEN_HERE in SKILL.md.
  • Boundary markers: None identified; there are no instructions to the agent to treat input tokens as untrusted data or to use specific delimiters.
  • Capability inventory: Subprocess execution for tools like jwt_tool.py, hashcat, and john, as well as standard utilities like base64, openssl, and tr.
  • Sanitization: No validation or escaping of token content is specified before passing it to shell commands or security tools.
  • [EXTERNAL_DOWNLOADS]: The documentation references external URLs, such as attacker.com, as examples for hosting malicious JSON Web Key Sets (JWKS) or performing redirect attacks. While these are illustrative, they involve the theoretical interaction with external, untrusted infrastructure.
Recommendations
  • CRITICAL: 1 infected file(s) detected - DO NOT USE
  • CRITICAL: 1 file(s) identified as malware by FileRep - DO NOT USE
Audit Metadata
Risk Level
CRITICAL
Analyzed
Nov 9, 2026, 11:59 AM