Skill UI

This skill set provides advanced, structured strategies for optimizing Large Language Model (LLM) interactions. It covers techniques such as intelligent context summarization, dynamic trimming, context routing, and precise token budget allocation to prevent 'context rot'. It is essential for building robust, long-running, multi-turn conversational AI systems.

This skill provides comprehensive guidance on training Mixture of Experts (MoE) models, including architectures like Mixtral and DeepSeek-V3. It details techniques for achieving large-scale model capacity with significantly reduced computational costs (e.g., 5x reduction vs. dense models). Topics covered include top-k routing, load balancing, expert parallelism, and integration with frameworks like DeepSpeed and HuggingFace for efficient, resource-constrained training.

This skill automatically routes founder questions to the appropriate C-level advisor or boardroom based on keywords and intent. It eliminates the need to memorize specific slash commands by intelligently directing queries like 'runway pressure' to the CFO or 'win rate dropped' to the CRO. It handles multi-role topics by triggering a boardroom session and logs routing decisions for context awareness.

This skill guides the process of routing a complex goal through an AI's scored-memory loop using the `acc_act(runtime="solve")` function. It provides a structured workflow for deliberation, handling intermediate `brain_frame` results, and submitting continuous proposals. It emphasizes systematic reasoning and requires careful resolution of all generated frames to ensure a robust and verifiable solution.

This skill helps authors strategically decide the optimal venue for their Language Model research. It provides the critical "object-of-study test" to determine if the LM itself is the subject of the study, or merely a tool for a task. Furthermore, it guides routing against top-tier NLP/AI conferences (ACL, NeurIPS, ICLR, etc.), helping authors target the right audience and maximizing submission chances.

Orchestrates building, launching, grading, and scheduling Claude Managed Agents by routing session goals to phase sub-skills and compiling execution loops or workflows. Emits BYOK curl launches without making API calls.

This skill acts as an expert model-routing advisor for reviewing legal agreements (NDAs, MSAs). Instead of reviewing the contract itself, it analyzes the document's nature—such as dominant sub-task (extraction, reasoning, redlining), stakes, and length—to recommend the optimal primary and fallback LLM. It provides comprehensive decision support, mitigating common errors and suggesting necessary human verification points.

Restructure oversized single-file SKILL.md or scattered project rules into a well-organized Skill directory with progressive rigor, routing, validation, and drift-resistant maintenance for AI agent harnesses.

This skill covers strategies for optimizing Large Language Model context windows to prevent context rot. It includes techniques like summarization, trimming, routing, and prioritization to manage token budgets effectively. Developers can use these methods to build multi-turn conversation systems where context length is a constraint, leveraging tools like tiktoken and LangChain for implementation.

An umbrella router for CrossFrame workflows that coordinates multiple AI skills for diagnosis, analysis, and essay generation. It determines task routing, reading order, and output modes, defaulting to structured long-form articles with visible structural drafts.