Train and deploy neural networks in distributed E2B sandboxes with Flow Nexus

Systematically deconstruct written content into verifiable claims, validate each using search/documentation, and facilitate informed discussion through structured interviewing.

Aggressive evidence-based audit to verify project claims match implementation reality

Create your Google Agent Development Kit skill in one prompt, then learn to improve it throughout the chapter

Create your LiveKit Agents skill from official documentation, then learn to improve it throughout the chapter

Create your Pipecat skill from official documentation, then learn to improve it throughout the chapter

Marketing and promotion specialist for Claude ecosystem technology - MCP servers, skills, plugins, and agents. Expert in community engagement, registry submissions, content marketing, and developer relations. Activate on 'promote MCP', 'share skill', 'market plugin', 'launch agent', 'developer marketing', 'MCP registry'. NOT for creating MCPs/skills (use agent-creator), general marketing (use content-marketer), or SEO optimization (use seo-visibility-expert).

Manage specification documents in .agent/specs/. Use when user provides requirements, acceptance criteria, or feature descriptions that need to be tracked and validated against implementation.

Test strategy, execution, and coverage analysis. Use when designing tests, running test suites, or analyzing test results beyond baseline checks.

Claude Code ecosystem expertise. Modules: CLI tool (setup, slash commands, MCP servers, hooks, plugins, CI/CD), extensibility (agents, skills, output styles creation), CLAUDE.md (project instructions, optimization). Actions: configure, troubleshoot, create, deploy, integrate, optimize Claude Code. Keywords: Claude Code, Anthropic, CLI tool, slash command, MCP server, Agent Skill, hook, plugin, CI/CD, enterprise, CLAUDE.md, agentic coding, agent, skill, output-style, SKILL.md, subagent, Task tool, project instructions, token optimization. Use when: learning Claude Code features, configuring settings, creating skills/agents/hooks, setting up MCP servers, troubleshooting issues, CI/CD integration, initializing or optimizing CLAUDE.md files.

Universal planning for technical and non-technical projects. Domains: software implementation, business, personal, creative, academic, events. Capabilities: feature planning, system architecture, goal setting, milestone planning, requirement breakdown, trade-off analysis, resource allocation, risk assessment. Actions: plan, architect, design, evaluate, breakdown, structure projects. Keywords: implementation plan, technical design, architecture, roadmap, project plan, strategy, goal setting, milestones, timeline, action plan, SMART goals, sprint planning, task breakdown, OKRs. Use when: planning features, designing architecture, creating roadmaps, setting goals, organizing projects, breaking down requirements.

Technical research methodology with YAGNI/KISS/DRY principles. Phases: scope definition, information gathering, analysis, synthesis, recommendation. Capabilities: technology evaluation, architecture analysis, best practices research, trade-off assessment, solution design. Actions: research, analyze, evaluate, compare, recommend technical solutions. Keywords: research, technology evaluation, best practices, architecture analysis, trade-offs, scalability, security, maintainability, YAGNI, KISS, DRY, technical analysis, solution design, competitive analysis, feasibility study. Use when: researching technologies, evaluating architectures, analyzing best practices, comparing solutions, assessing technical trade-offs, planning scalable/secure systems.

Structured reflective problem-solving methodology. Process: decompose, analyze, hypothesize, verify, revise. Capabilities: complex problem decomposition, adaptive planning, course correction, hypothesis verification, multi-step analysis. Actions: decompose, analyze, plan, revise, verify solutions step-by-step. Keywords: sequential thinking, problem decomposition, multi-step analysis, hypothesis verification, adaptive planning, course correction, reflective thinking, step-by-step, thought sequence, dynamic adjustment, unclear scope, complex problem, structured analysis. Use when: decomposing complex problems, planning with revision capability, analyzing unclear scope, verifying hypotheses, needing course correction, solving multi-step problems.

**Core Principle:** Orchestrate skills dynamically. Analyze the task, discover available skills, build the right chain, explain the reasoning, execute step-by-step with confirmation.

> **도메인:** devian-upm-samples

To maintain the integrity of audit logs, we must ensure that agents use their real identities in worklogs, rather than blindly copying "codex" from templates.

[01] META. Сканирует доступные skills, создает план выполнения и идет шаг за шагом с подтверждением каждого этапа. Triggers on complex tasks, multi-step work, or when structured execution is needed.

> **도메인:** devian-upm-samples

The skill is script-heavy and evolving fast. When Python dependencies are missing, agents currently discover failures late (mid-command) and then waste tokens installing packages ad-hoc. This also red

[02] META. Create new skills when existing ones don't cover the task. Analyze unique requirements, build framework (Frame → Research → Plan → Execute), integrate risks, and declare new skill. Use when facing novel problems that existing skills can't address.

Transforms conversations and discussions into structured documentation pages in Notion. Captures insights, decisions, and knowledge from chat context, formats appropriately, and saves to wikis or databases with proper organization and linking for easy discovery.

Establish clarity before starting work. Use when beginning any significant task, when input is vague or stream-of-consciousness, or when requirements seem unclear. Handles messy voice input efficiently. This is the first system in the 5-system framework.

A personal knowledge base for **building understanding that compounds over time**. Not a note dump - a structured system for capturing knowledge you can actually retrieve and use.

Turn ideas into clear, buildable specs for AI tools or stakeholder review. Use when starting features, planning projects, or when AI keeps building the wrong thing. Creates Quick Feature Specs (10-15 min) for immediate AI builds or Full Project Scopes (1-2 hours) for budget planning and contractor estimates.

The Twelve-Factor App methodology for building scalable, maintainable cloud-native applications. Use when designing backend services, APIs, microservices, or any software-as-a-service application. Triggers on deployment patterns, configuration management, process architecture, logging, and infrastructure decisions.

This skill should be used when users need to infer chromatin states from histone modification ChIP-seq data using chromHMM. It provides workflows for chromatin state segmentation, model training, state annotation.

Guide for adding new blockchain chains to OneKey. Use when implementing new chain support, adding blockchain protocols, or understanding chain architecture. Triggers on chain, blockchain, protocol, network, coin, token, add chain, new chain.

Adds new WebSocket event subscriptions to OneKey. Use when implementing new socket events, handling server push messages, or adding real-time data subscriptions. Socket, WebSocket, event, subscription, push, real-time.

Interactive guide for integrating new DeFi modules or protocols into OneKey. Use when adding new DeFi features like staking protocols, lending markets, or entirely new DeFi modules. Triggers on DeFi, protocol, integration, Earn, Borrow, staking, lending, supply, borrow, withdraw, repay, claim, new module, Pendle, Aave, Compound.

Git workflow and conventions for OneKey development. Use when creating branches, committing code, or creating PRs. Triggers on git, branch, commit, PR, pull request, merge, workflow.

Creates a new Claude Code Skill following best practices. Use when the user wants to create a new skill, add a skill, or asks about writing skills for Claude Code. Fetches latest documentation before generating skill content. New skill. Create a skill.

Documents minimum SDK/OS version requirements for all OneKey platforms. Use when checking platform compatibility, understanding deployment targets, verifying version requirements, or when user asks if their device can run the project. Triggers on minimum version, SDK version, API level, deployment target, platform requirements, iOS version, Android version, Chrome version, Electron version, can I run, environment check, device compatibility, check environment.

1password-secrets

Create and manage project-specific design systems for SaaS applications. Use when starting a new app (CREATE mode), updating an existing design system (MODIFY mode), or building features that need to follow established patterns (WORK WITHIN mode). Outputs to instructions/design-system.md as the single source of truth for all UX/UI decisions.

Identify friction points, bottlenecks, bugs, and technical debt. Use for audits, debugging sessions, when something feels wrong, or before major work to surface hidden issues. This is the second system in the 5-system framework.

**Core Principle:** Achieve complete clarity, not just confirm understanding. Know the WHAT, WHY, and HOW of success before acting.

>

Generated skill from request: pattern emergence detector

Generated skill from request: trinity auto-boot validator

[21] UNDERSTAND. Deep research workflow for this project using 150% scope (100% core + 50% boundary), evidence-based reasoning, and structured investigation notes. Use when the task requires investigation, root-cause analysis, or mapping unknown areas. Always maintain a research log file that captures findings, hypotheses, and next branches; use web.run when external verification is needed.

**Core Principle:** Research deeply from ALL sources — internal context AND external knowledge. Don't limit yourself to one world.

[23] UNDERSTAND. Consult external AI models when internal sources are exhausted. Build quality prompts using Prompt150 formula (Context + Query + Method + Style). Use when Loop150 exhausts internal sources, need real-world precedents, confidence <75%, or require reasoning from specialized AI models.

Prepares meeting materials by gathering context from Notion, enriching with Claude research, and creating both an internal pre-read and external agenda saved to Notion. Helps you arrive prepared with comprehensive background and structured meeting docs.

Creates high-quality content (blog posts, tweets, newsletters, documentation) that matches the user's writing style and voice. Performs web research to find citations and supporting evidence. Use when user requests blog posts, marketing content, newsletters, tweets, or any written content that should sound authentic and be well-researched.

**Core Principle:** Understand the ripple effects. See how changes in one part affect the whole system.

ALWAYS use this skill when user needs ANY API functionality (AI models, image generation, video, audio, text processing, etc.). Automatically search 302.AI's 1400+ APIs and generate integration code. Use proactively whenever APIs or AI capabilities are mentioned.

[32] ANALYZE. Ensure every critical claim has verifiable evidence with confidence levels. Each fact must have source + confidence percentage. If confidence <85%, enter Loop150 to find more sources. Use for critical decisions, factual claims, legal/compliance work, or any situation where unverified claims are dangerous.

This skill annotates chromatin loops, including enhancer/promoter assignments, CTCF-peak overlap. It automatically constructs enhancer and promoter sets when missing and outputs standardized loop categories.

Deep knowledge expert for the 33GOD agentic pipeline system, understands component relationships and suggests feature implementations based on actual codebase state

Create your OpenAI Agents SDK skill in one prompt, then learn to improve it throughout the chapter