LLM & Agents

Real-time asciinema recording backup to GitHub orphan branch with idle-based chunking and brotli archival. TRIGGERS - streaming backup, recording backup, asciinema backup, continuous recording, session backup, orphan branch recording, zstd streaming, chunked recording, real-time backup, github recording storage.

Optimize CLAUDE.md files and Skills for Claude Code CLI. Use when reviewing, creating, or improving system prompts, CLAUDE.md configurations, or Skill files. Transforms negative instructions into positive patterns following Anthropic's official best practices.

Orchestrate baselayer subagents for complex tasks. Defines available agents, their skills, and workflows for multi-agent scenarios. Load when coordinating work across agents, delegating tasks, or deciding which agent handles what.

Launches specialized Claude agents for targeted tasks. Analyzes requirements, selects appropriate agent, and executes with optimized configuration.

State-driven interactive animations with built-in state machines. Use when animations must REACT to user input (hover, click, drag), have multiple states/transitions, or respond to data values (progress bars, counters). Ideal for animated buttons, toggles, checkboxes, characters. For simple play/loop animations use Lottie instead.

This skill should be used when users need to generate ideas, explore creative solutions, or systematically brainstorm approaches to problems. Use when users request help with ideation, content planning, product features, marketing campaigns, strategic planning, creative writing, or any task requiring structured idea generation. The skill provides 30+ research-validated prompt patterns across 14 categories with exact templates, success metrics, and domain-specific applications.

Explain what this skill does and when Claude should use it. Include trigger keywords and use cases.

Validate datasets for Unsloth fine-tuning. Use when the user wants to check a dataset, analyze tokens, calculate Chinchilla optimality, or prepare data for training.

Analyze a specific directory (e.g. frontend/, backend/, e2e/) and generate .claude/rules/ markdown files for it. Use when asked to create rules, analyze a folder for Claude Code, or set up domain memory for a specific part of the codebase.

Provides structured error handling protocol for all agent operations. Ensures users receive clear error reports, troubleshooting guidance, and actionable next steps. Use when any tool operation fails, encounters an error, or produces unexpected results.

Patterns for delegating UI work to specialized agents. Covers synthesis-master vs specialized agents, multi-agent UI generation workflows, and orchestration strategies for complex UI tasks.

全范围召回 - 从 Project CLAUDE.md、User CLAUDE.md 和 Memory MCP 查询历史知识

Guide for creating effective skills. Use when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.

Configure scan-mcp MCP server in Claude Code's global configuration. Use when user wants to setup, configure, initialize, enable, or install the scan-mcp MCP server. Runs preflight checks for prerequisites (Node 22+, SANE tools, tiffcp), helps install missing dependencies, and adds server configuration with user-specified INBOX_DIR.

Telegram bot production management, monitoring, and troubleshooting. Use when user mentions telegram bot, claude-orchestrator, bot status, bot restart, or bot monitoring.

AshAi extension guidelines for integrating AI capabilities with Ash Framework. Use when implementing vectorization/embeddings, exposing Ash actions as LLM tools, creating prompt-backed actions, or setting up MCP servers. Covers semantic search, LangChain integration, and structured outputs.

Manages Claude configuration files for both Claude Desktop and Claude Code projects. Handles MCP server configuration, environment variables, project settings, and developer options. Use when configuring Claude Desktop with MCP servers, setting up Claude Code project settings, managing environment variables for Claude, enabling debugging features, or troubleshooting Claude configuration issues.

Troubleshoot Claude Code session issues. Use when encountering "No conversations found" errors, missing sessions, or session file corruption problems.

Embrace vibe coding for rapid UI exploration. Covers when to iterate vs refine, ephemeral app patterns, and the art of fast, disposable prototyping with AI-assisted development.

Use when creating or developing, before writing code or implementation plans - refines rough ideas into fully-formed designs through collaborative questioning, alternative exploration, and incremental validation. Don't use during clear 'mechanical' processes