Skill UI

A systematic framework for launching and validating new products and business models. It guides users through the Build-Measure-Learn loop, emphasizing validated learning over vanity metrics, and structuring Minimum Viable Product (MVP) experiments to test core assumptions and determine if a pivot or perseverance is necessary.

Guides product teams through hypothesis writing, metric definition, sample-size planning, ICE prioritization, and statistical interpretation so experiments stay practical, defensible, and aligned with business decision thresholds.

This skill explains complex concepts using the Socratic method. Instead of delivering direct answers or textbook definitions, it engages the user in a reflective, conversational dialogue. By guiding the user through thought experiments, small observations, and simple analogies, the assistant helps the user construct understanding step-by-step, facilitating deeper clarity through active reasoning. Ideal for understanding theories, abstract ideas, or system mechanics.

A structured methodology for de-risking product ideas and validating market opportunities. This framework guides teams through defining measurable outcomes, building Opportunity Solution Trees (OST), mapping critical assumptions (Desirability, Viability, Feasibility), and planning rigorous experiments (prototyping, interviews, usability tests) before committing to full development resources. Ideal for discovery sprints and pre-build validation.

Starts a recurring autoresearch experiment loop at user-selected intervals (10 minutes to monthly), auto-schedules CronCreate jobs, records experiment metadata, and offers commands to monitor progress or stop the loop.

Resumes paused experiments by listing branches, loading config/strategy/results, summarizing past runs, and prompting the next action so you can continue iterative tuning without losing context.

Interactive command that gathers domain, target file, evaluation command, metric, direction, evaluator, and scope to initiate a new autoresearch experiment and list existing experiments or evaluators.

This skill orchestrates end-to-end autonomous AI research projects using a sophisticated two-loop architecture. It manages the entire research lifecycle, from initial literature review and hypothesis formation to running constrained experiments and synthesizing findings into publishable papers. It acts as a project manager, autonomously driving the process while structuring the workspace and presenting progress to human users.

Processes machine learning experiment outputs to build comparison tables, compute statistics, and surface insights with recommendations for follow-up trials whenever users ask to analyze results or make comparisons.

This skill transforms a raw research proposal or method idea into a rigorous, paper-ready experimental roadmap. It guides the user through defining core claims, identifying necessary ablations (e.g., novelty, simplicity, frontier necessity), and structuring experiments into actionable, phased stages (Sanity, Baseline, Main, Decision). Ideal for advanced academic research in AI/LLM/VLM fields that requires strong validation and systematic reporting.

Internal Feishu/Lark notification utility that other skills call to push event cards or run interactive dialogs via webhook or bridge, ensuring safe fire-and-forget delivery whenever checkpoints, experiments, reviews, or pipelines change state.

This comprehensive skill transforms a broad research direction into a structured, actionable pipeline for academic discovery. It first performs a deep literature survey (local and web) to map the current state-of-the-art and identify structural gaps. It then utilizes advanced LLMs to generate, hypothesize, and rank 8-12 concrete research ideas, providing hypotheses, minimum viable experiments, and estimated effort, making them highly publishable.