Skill UI

A meta-skill designed to monitor long-running processes, such as model fine-tunes or CI builds, executed within a detached tmux session. It captures output, automatically uses a sub-agent to diagnose failures, proposes a fix, and records the complete debugging process to memory. Ideal for intermittent failure debugging and continuous integration pipelines.

A comprehensive minimal terminal coding harness designed for building, extending, and integrating advanced AI capabilities directly into CLI environments. It supports deep integration via SDK (Node/TypeScript), RPC, and JSON event streams. Developers can create custom skills, providers, TUI components, and utilize advanced features like web search, video understanding, and agent orchestration.

This skill allows users to mine their extensive coding history across various AI agents (Claude, Codex, Gemini, etc.) into a persistent, structured corpus. It surfaces reproducible workflows as evidence and interactively helps the user forge these past practices into reusable, installed skills across their agent ecosystem. It is designed to synthesize judgment from raw transcripts into actionable, portable skills.

Max is a specialized agent designed for safe, non-behavioral code improvement. It systematically refactors existing, working code by optimizing time complexity (Big-O), abstracting duplicated logic, removing dead code, and enhancing readability. It operates under a strict constraint: no change may alter the original functionality, requiring all refactored code to pass existing test suites flawlessly.

Rex is a foundational agent designed to transform vague, high-level user intent into precise, unambiguous, and structured technical specifications. It acts as a requirements analyst, asking clarifying questions, identifying edge cases, and defining scope using frameworks like MoSCoW and User Stories. It ensures that all subsequent development stages (planning, architecture, implementation) work from a single, agreed-upon artifact, minimizing guesswork and scope creep.

This skill designs and audits the core engineering harness required for reliable AI agent collaboration within a codebase. It establishes critical infrastructure elements such as change tracking, linting rules, CI/CD gates, and standardized documentation, ensuring the repository remains the single source of truth for all intelligent agents. Ideal for setting up complex development workflows.

This library manages the entire lifecycle of bounded AI agent feedback loops. It allows users to find existing workflows from a catalog, adapt established loops by modifying parameters or tools, or design entirely new, structured iterative processes through a guided interview. It emphasizes safety by enforcing explicit checks, stop rules, guardrails, and measurable terminal states, preventing endless, uncontrolled autonomy.