Skill UI

Offensive security skill for hunting insecure deserialization flaws in Java, PHP, Python, .NET and Ruby. Covers ysoserial gadget chains, PHP object injection, pickle payloads, ViewState and JNDI/Log4Shell, with strict authorization gates.

This skill covers identifying and exploiting HTTP request smuggling vulnerabilities (CL.TE, TE.CL, H2.CL, H2.TE). It includes fingerprinting front-end servers, crafting payloads for cache poisoning, and harvesting credentials via backend desyncs. Requires authorization and defensive guidance protocols.

Methodology for hunting open redirect flaws in web apps: parameter discovery, bypass payloads, DOM-based sinks, OAuth auth-code theft chains and SSRF escalation. Docs-only, for authorized bug bounty and pentest use.

Detects and exploits server-side template injection (SSTI) across Jinja2, Twig, Freemarker, ERB, Spring, Velocity, Mako, Thymeleaf, and Smarty, offering detection payloads, RCE escalation paths, and authorized testing guidance.

Hunts reflected, stored and blind XSS during authorized web security assessments: reflection verification, unique canary payloads, context-based injection and out-of-band confirmation.

A workflow for converting known binary vulnerabilities into working exploits. Covers stack/heap/kernel pwn techniques using pwntools, libc-database, and ROP. Suitable for porting CTF solutions to authorized remote targets or developing RCE payloads for Linux kernel drivers.

RouterBase acts as a unified, OpenAI-compatible API gateway, allowing applications to route requests across diverse large language models (LLMs) like GPT, Claude, and Gemini. It helps developers migrate existing OpenAI SDK calls, implement sophisticated primary/fallback model selection based on cost, latency, and quality, and manage multimodal payloads (image, audio, video) from a single, standardized endpoint.

A security payload collection containing XSS, SSRF, and SQL injection payloads, CSP/WAF bypass techniques, wordlists, and submission rules for authorized penetration testing.

This comprehensive, zero-trust security auditor scans third-party AI agent skills before installation. It performs a rigorous 6-phase review process, analyzing surface patterns, script integrity, permissions, and supply chain risks. It is designed to detect malicious payloads, credential theft, instruction overrides, and prompt injection attacks, ensuring secure AI development and deployment.

Resolve the newest complete numerical weather prediction cycle and forecast objects across provider mirrors without downloading full payloads. This skill probes exact objects or narrow prefixes, verifies request-specific completion contracts, and returns the newest qualifying initialization with source metadata.

Creates disposable private webhook, email, and DNS inboxes via MCP so agents can wait for asynchronous callbacks, inspect received payloads, validate integration test events, and clean up safely.