hunt-idor

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill includes examples and instructions that embed session cookies and bearer tokens directly into curl/ffuf/Intruder commands (e.g., "Cookie: session=USER_A_SESSION", "Authorization: Bearer USER_B_TOKEN") and tells the user to "replace session cookie/token" with credentials, which requires the LLM to output secret values verbatim and therefore poses a high exfiltration risk.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This document is an offensive IDOR exploitation playbook that explicitly instructs how to access, enumerate, exfiltrate, and take over other users' accounts and data, including examples that send stolen tokens/cookies to attacker-controlled endpoints.

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

• Direct money access detected (high risk: 1.00). The skill explicitly documents and instructs calling financial state-change endpoints — e.g., redirecting charges via voucher/policy modification and issuing refunds with a POST to /api/orders/{order_id}/refund — which are concrete instructions to move money. It also references payment-related mutations and a Stripe-related update in real examples. These are specific financial-execution primitives, not generic tooling, so the skill grants direct financial execution capability.