The Bounty Security Pattern Master Library is a skill that loads a machine-readable JSON dataset of 399 structured vulnerability patterns into an AI agent’s context, enabling targeted security audits and bug bounty research. The library covers classic web vulnerabilities mapped to CWE identifiers as well as emerging DeFi-specific attack categories including oracle manipulation paths, withdrawal accounting flaws, and cross-chain bridge redemption issues. A dedicated subset labeled “Monetrix” patterns addresses Hyperliquid-style and cross-chain protocol vulnerabilities that are not well represented in general-purpose security references.
For each pattern, the library provides exploitation hypotheses, root cause mappings, and false-positive checklists, which helps an agent move from generic code review toward reasoning about specific, exploitable scenarios. Agents can use the structured data to generate formatted audit summaries ready for triage or bounty platform submission, reducing manual effort in translating a finding into a reportable form.
The skill is designed for developers and security researchers who already have code or a protocol under review and want an agent to apply domain-specific knowledge rather than relying on a base LLM’s general security reasoning. Because no tools are exposed and no environment variables are required, the skill functions purely as a knowledge layer injected into the agent’s working context. It does not perform live scanning, connect to external APIs, or interact with any blockchain node.