Skill UI

Speeds up LLM inference by combining speculative decoding with Medusa multiple heads and lookahead techniques so chatbots, code assistants, and other real-time workloads can run with 1.5-3.6× lower latency while preserving target-model quality.

An adaptive and persistent coaching system designed for the entire job search lifecycle. It covers critical stages including Job Description (JD) decoding, resume optimization, mock interviews (behavioral, technical, case), transcript analysis, storybank building, and comprehensive salary negotiation support. The system tracks your progress and provides tailored coaching plans across 23 commands.

This skill provides methods for reliably extracting typed, validated, and structured data from Large Language Models (LLMs), moving beyond simple text parsing. It covers best practices across major APIs (OpenAI's response_format, Anthropic's tool_use, Google's responseSchema) and local models. Key workflows include defining schemas (Pydantic/Zod), implementing constrained decoding, and building robust retry loops for production-grade data pipelines.

A comprehensive, multi-step workflow for analyzing and deobfuscating highly complex and malicious JavaScript code. It guides users through extracting scripts from phishing pages and web skimmers, identifying obfuscation techniques (Base64, Unicode escapes, eval chains), neutralizing anti-analysis features, and progressively decoding string layers to reveal the original malicious payload for thorough threat intelligence and malware analysis.

Fetches a Gmail message by ID and extracts its body or headers, with options for text or JSON output, HTML decoding, and dry-run previews to streamline email inspection in scripts.

This technique guides users in solving trivia or puzzle questions where the target band's country is not explicitly named. Instead, the country is hinted at through indirect geographic clues such as land area ranges, EU membership dates, or national flag descriptions. It emphasizes systematic deduction, requiring the user to treat the clue as a multi-candidate puzzle and test all viable countries before refining the search to the specific band name.

This comprehensive guide serves as a complete error decoder for the OneNote Graph API. It analyzes major HTTP status codes (400, 403, 404, 429, 500, 502, 507) and provides root cause analysis, diagnostic steps, and code fixes for common issues. It specifically addresses critical failure modes like deprecated authentication methods and silent 200 OK failures, helping developers debug complex API integrations in production.

This comprehensive playbook provides expert techniques for classical cipher identification and breaking. It covers fundamental methods like frequency analysis, Kasiski examination, and Index of Coincidence (IC), enabling the decryption of monoalphabetic, polyalphabetic, and transposition ciphers (Vigenere, Caesar, Playfair, etc.). It also guides the decoding of various formats (Base64, Hex, Morse).

A comprehensive guide detailing the standards required for academic reproducibility in NLP and LLM research. It covers the 'Responsible NLP checklist' requirements, including detailed reporting for hyperparameters, model versions, exact prompts, decoding parameters, data contamination auditing, and variance reporting to ensure claims are auditable and robust.

A comprehensive guide detailing best practices for designing and auditing empirical evaluations of Large Language Models (LLMs). It addresses critical pitfalls such as data contamination, baseline unfairness, model version drift, and decoding parameter sensitivity, ensuring that published results are robust, reproducible, and scientifically sound for academic publication.

A comprehensive guide detailing how to structure and populate the supplementary material section of technical research papers, especially in NLP and Machine Learning. It distinguishes between load-bearing elements (main text) and audit trails (appendix), covering prompt verbatim, decoding configurations, full result grids, and code anonymization protocols to ensure high reproducibility and academic rigor.

This guide provides a deep dive into the rigorous FOCS (Foundations of Computer Science) review process. It models the entire academic pipeline—from initial assignment and expert consultation to committee discussion and final decision—allowing authors to predict how their work will be evaluated based on correctness, novelty, and overall field contribution in a double-blind setting.