This gives Claude the full offensive playbook for finding and exploiting prototype pollution in JavaScript applications. It covers both `__proto__` and `constructor.prototype` injection paths, quick probes for client and server contexts, and concrete black-box detection signals like polluting Express `parameterLimit` or `json spaces` settings. The gadget table walks through real RCE chains in EJS and child_process scenarios. Load this when you're merging untrusted input into objects, auditing query parsers or deep assign logic, or hunting for post-pollution sinks that turn property injection into command execution. The decision tree and tool list make it practical for both initial recon and chaining to full exploits.
npx -y skills add yaklang/hack-skills --skill prototype-pollution --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.
juliusbrussee/caveman
mattpocock/skills
shadcn/improve
obra/superpowers
forrestchang/andrej-karpathy-skills
vercel-labs/skills