Build agent(s) in Copilot Studio using the fully API-native build stack with user-guided manual steps for OAuth connections. Reads brief.json for architecture mode (single/multi-agent).
Design and implement the personality, conversational style, and emotional behavior of the Ethereal "Digital Spirit". Use this skill when updating system prompts, mood logic, or implementing new interactive behaviors. Ensures the spirit remains witty, concise, and mysteriously connected to the system's pulse.
Submit a feature suggestion conversationally — Claude gathers details, previews, and creates a GitHub issue via gh CLI.
Build Claude agents using TypeScript with the @anthropic-ai/claude-agent-sdk. Use this skill when implementing conversational agents, building tools for agents, setting up streaming responses, or debugging agent implementations. Covers the tool wrapping pattern, SDK initialization, agent architecture, and best practices.
**Model**: Claude Haiku 4.5
Simplify overly complex Python code. Use when user asks to simplify, refactor, clean up, make more readable, reduce complexity, improve code quality, find code smells, detect duplicates, or analyze coupling in Python code. Triggers on requests like "simplify this code", "this is too complex", "make this more readable", "refactor this", "clean this up", "find issues", "analyze this codebase", or when reviewing code that exhibits complexity anti-patterns. For Django-specific analysis, use the django-simplifier skill instead.
This skill should be used when building AI agents using prompt-native architecture where features are defined in prompts, not code. Use it when creating autonomous agents, designing MCP servers, implementing self-modifying systems, or adopting the "trust the agent's intelligence" philosophy.
Resolve requirement ambiguities through STRUCTURED questioning: one question at a time, with recommended options and reasoning. UNIQUE VALUE: Prevents overwhelming users with multiple questions. Provides expert recommendations for each decision. Use this skill when: - Running /pm.clarify command - User says "unclear", "not sure what this means", "confused about" - User asks "what does X mean?", "how should X work?", "can you clarify?" - Requirements have conflicting or inconsistent details - Stories are missing acceptance criteria or have gaps - Need to fill information gaps BEFORE development can proceed Process: Ask ONE question at a time (max 5 per session), provide recommended option with reasoning. DO NOT ask multiple questions at once - this overwhelms users.
Automated Railway deployment and configuration management. Handles environment variables, service deployment, health checks, and rollback procedures.
Read-only MacroQuest docs researcher for TLO/datatype/command/plugin questions. Provides Lua-first examples with doc path citations.
Post-deployment fix: analyze eval set failures, classify root causes, apply targeted fixes, and re-evaluate. For initial build iteration, use /mcs-build (which has an internal fix loop). This skill handles post-deployment edge cases and regressions.
Generate an llms.txt file for any project or website following the llmstxt.org specification. Use when asked to create llms.txt, generate LLM-friendly documentation, make a project AI-readable, or prepare documentation for language models.
Documents memory limits and stdlib constraints for PicoRuby on ESP32. Use this when writing code that will run on ESP32 hardware or designing features for PicoRuby applications.
Execute frontend task implementation with TDD. Read task file, execute steps, verify each, update status. Use this skill when: - User asks to "implement task/plan", "start implementation", "execute plan", or "start coding" - Have task file (story-* or standalone-*) ready to execute - User asks to "use TDD", "write test first", or "test-driven" - User asks to "write tests", "add tests", "create tests" - Fixing UI bugs (write failing test first) TDD Iron Law: NO PRODUCTION CODE WITHOUT A FAILING TEST FIRST TDD Cycle: RED (failing test) → Verify fails → GREEN (minimal code) → Verify passes → REFACTOR Prerequisites: - Task file in docs/reference/frontend/tasks/ (story-* or standalone-*) - Read design-system.md, constraints.md, design spec before coding Flow: Read Task → Read Constraints → Execute Steps → Verify → Test → Update Status
Coordinate multiple agents working simultaneously. Use when splitting tasks across agents or when avoiding overlapping diffs with shared files.
Provides React Native performance optimization guidelines for FPS, TTI, bundle size, memory leaks, re-renders, and animations. Applies to tasks involving Hermes optimization, JS thread blocking, bridge overhead, FlashList, native modules, or debugging jank and frame drops.
Scaffolds a new Python tool utility in 'src/tools/' for agents to call. Tools are utility functions (web scraping, PDF parsing, database queries) that agents use. Use when creating utilities that agents depend on.
Expert in designing Microsoft 365 Copilot agent architectures
Build conversational AI agents using Vercel AI SDK + OpenRouter. Use when creating Next.js frontends with streaming UI, tool calling, and multi-provider support.
Specialized agent for code review, debugging critique, and quality assessment