jwt-oauth-token-attacks

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill includes many CLI examples and tool invocations that require inserting real JWTs, keys, or secrets verbatim (e.g., JWT_TOKEN, JWT_TOKEN_HERE, public_key.pem used as an HMAC secret, hashcat/john invocations), so an agent following it would be expected to handle and output secret values directly, creating exfiltration risk.

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 1.00). Yes — several URLs point to attacker-controlled hosts (attacker.com/malicious-jwks.json, attacker redirection patterns, userinfo-host URLs, localhost redirect and open-redirects) that are commonly used to fetch malicious keys, exfiltrate tokens, or redirect users to attacker-hosted payloads, so they are suspicious.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This document is an explicit offensive playbook providing step-by-step techniques and tooling to forge JWTs, host malicious JWKS, brute-force secrets, and steal OAuth tokens—clearly enabling credential theft, token exfiltration, and account takeover.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill includes a runtime attacker-controlled JWKS URL (https://attacker.com/malicious-jwks.json) used in examples and an automation command to host/fetch keys that would directly influence token verification at runtime.