Skill UI

An advanced cognitive toolkit for AI agents, exposing four specialized modes: reasoning, code, anti-deception, and memory. Instead of generating raw thoughts, it forces the agent to ingest a structured 'scaffold' (including failure patterns, procedures, and falsification tests) before forming a final output. This mitigates common AI failure modes like hallucination, decay, and sycophancy, significantly enhancing reliability in complex tasks.

This system simulates a full C-suite executive team, featuring 8 specialized persona agents (CFO, CMO, etc.) and 17 deep-dive slash commands. It facilitates structured decision-making, strategic planning (e.g., sprint pipelines), and cross-functional deliberation through simulated office hours and boardrooms. Ideal for founders needing instant, multi-perspective executive guidance.

This command initiates a structured founder interview to capture critical, durable facts about the company. It covers basics, business models, financials, and strategic priorities across 12 key areas. The output, `company-context.md`, serves as the single source of truth for all AI agents and advisors, ensuring all decision-making is based on accurate, comprehensive company context rather than assumptions.

Expert skill for building Model Context Protocol (MCP) servers and tools from scratch. Covers the full development lifecycle—from defining tool schemas and resources to implementation, testing, and publishing to registries. Supports robust patterns using TypeScript and Python for integrating custom capabilities into AI agents.

This skill simulates a comprehensive international journal peer review process by deploying five specialized agents (EIC, 3 peer reviewers, and a Devil's Advocate). It reviews papers from diverse angles—including methodology rigor, domain expertise, cross-disciplinary views, and core arguments—to produce a structured editorial decision and detailed revision roadmap. Ideal for academic researchers seeking high-quality, multi-faceted feedback.