What problem does it solve? Broken object-level authorization (IDOR/BOLA) is one of the most common and damaging API flaws, yet it is hard to detect with generic scanners because it requires comparing responses across users, roles, and tenants. This Skill provides a systematic methodology for finding horizontal/vertical privilege escalation, predictable identifier abuse, and cross-tenant isolation failures before attackers do. ## Core Features & Use Cases - Comprehensive IDOR attack surface coverage: Tests horizontal/vertical privilege escalation (BOLA/BFLA), mass assignment, prototype pollution, GraphQL node queries, WebSocket channel subscription, and JWT claim tampering across REST, GraphQL, and gRPC endpoints. - Predictable identifier reverse engineering: Decodes UUIDv1 timestamps, Snowflake IDs, ULIDs, Hashids, and Base62 short codes, plus entropy analysis to determine whether "random" IDs are actually enumerable. - Multi-tenant and AI platform testing: Covers cross-tenant isolation failures (X-Tenant-ID switching, JWT tenant claims, cache key leakage) and LLM platform object authorization (flows, threads, credentials, knowledge bases) with real CVE case studies. - Automated BOLA detection methodology: Multi-account session comparison using Autorize, AuthMatrix, and AutoRepeater with response body diffing rather than status-code-only checks. - Use Case: During an authorized API penetration test, configure two test accounts, replay all captured requests with the low-privilege session, and use the response-diffing checklist to identify endpoints leaking other users' order, invoice, or profile data. ## Quick Start Ask the AI to test the target application's user profile and order API endpoints for horizontal and vertical IDOR vulnerabilities using two authorized test accounts.