自然語言處理

Framework for starting new daily chat with gentle context loading from previous day's summary. Use when user says "good morning", starts new chat with minimal prompt, or explicitly requests morning brief. Loads yesterday's summary and generates scannable brief. Optimized for low cognitive load during morning wakeup. Extend with personal metrics and protocols.

AI-powered image generation and editing using Google Gemini, Google Imagen, and OpenAI models. Generate images from text descriptions, edit existing images, create logos/stickers, apply style transfers, and produce product mockups. Use this skill when the user requests: - Image generation from text descriptions - Image editing or modifications - Logos, stickers, or graphic design assets - Product mockups or visualizations - Style transfers or artistic effects - Iterative image refinement Available models: - Google Gemini: gemini-2.5-flash-image (Nano Banana), gemini-3-pro-image-preview (Nano Banana Pro) - Google Imagen: imagen-4.0-generate-001, imagen-4.0-ultra-generate-001, imagen-4.0-fast-generate-001 - OpenAI: gpt-image-1.5 (recommended), gpt-image-1, dall-e-3, dall-e-2 Inspired by: https://github.com/EveryInc/every-marketplace/tree/main/plugins/compounding-engineering/skills/gemini-imagegen

Always use this skill to provide context on the codebase. Automatically initializes and references the Speculate documentation framework (coding rules, shortcuts, templates, and project specs) to ensure agents understand project structure, development workflows, and coding standards. Use for every task to maintain context and code quality.

READ-ONLY quick overview of GitHub Elements workflow state. Shows active threads, phase distribution, and workflow health at a glance. USE THIS SKILL WHEN: - User asks "what's the status" or "show me the status" - User asks "what threads are active" or "what am I working on" - User asks "show me the workflow state" or "what's happening" - User wants a quick overview before starting work - Starting a session and need context DO NOT USE THIS SKILL WHEN: - User wants to CLAIM an issue (use ghe-claim) - User wants to POST a checkpoint (use ghe-checkpoint) - User wants to TRANSITION phases (use ghe-transition) - User wants DETAILED metrics/health reports (use ghe-report) EXAMPLES: <example> Context: User starting a session wants to see current state user: "What's the github elements status?" assistant: "I'll use ghe-status to show you the current workflow state" </example> <example> Context: User wants to know what work is available user: "Show me what threads are active" assistant: "I'll use ghe-status to display active and available threads" </example> <example> Context: Quick check before doing work user: "What am I currently working on?" assistant: "I'll use ghe-status to find your in-progress threads" </example>

Prepare meeting materials with Notion context and Codex research; use when gathering context, drafting agendas/pre-reads, and tailoring materials to attendees.

MCP-powered domain research for requirements elicitation. Uses perplexity, context7, firecrawl, and other MCP servers to research domain knowledge, best practices, and industry requirements.

Retrieve project-specific insights from file-based memory. Use when working on features, encountering domain-specific questions, or user says "--project-recall" or "--recall" (you decide which scope, may use both). Skip for routine tasks or universal pattern questions (use coder-memory-recall). This Skill MUST be executed using Task tool with subagent_type="general-purpose". Runs in separate context to avoid polluting main conversation.

This skill should be used when the user asks to "add AI chat", "implement streaming UI", "use useChat hook", "add AI completion", "implement useCompletion", "create conversational interface", "add streaming text", "implement tool calling UI", "create generative UI", "add AI-powered features", "implement StreamableUI", or mentions AI SDK, streaming responses, chat interfaces, or AI-generated content in React/Next.js applications.

MCP (Model Context Protocol) server patterns with categorical tool composition and typed context protocols. Use when building MCP servers with categorical composition patterns, designing tool interfaces using functors and natural transformations, implementing typed context management, or creating composable AI tool ecosystems with categorical guarantees.

Executes tdx CLI commands for Treasure Data. Covers `tdx databases`, `tdx tables`, `tdx query`, `tdx auth setup`, context management with profiles/sessions, and output formats (JSON/TSV/table). Use when users need tdx command syntax, authentication setup, database/table exploration, or query execution.

Match file paths against steering file glob patterns to determine applicable steering guidance. Use when orchestrator needs to inject context-aware guidance based on files being modified.

CHANGE phases in the GitHub Elements workflow. Closes current thread and opens next phase thread. This is a PHASE CHANGE operation, not just saving progress. USE THIS SKILL WHEN: - User says "transition to TEST" or "move to REVIEW" - User says "I'm done with DEV, start TEST" or "DEV complete" - User says "advance to next phase" or "close this phase" - User says "demote to DEV" (from TEST or REVIEW) - User says "tests pass, ready for review" - User says "merge this" (after REVIEW PASS) - User explicitly wants to END current phase and BEGIN next DO NOT USE THIS SKILL WHEN: - User just wants to SEE status (use ghe-status) - User wants to START work on new issue (use ghe-claim) - User wants to SAVE progress but CONTINUE working (use ghe-checkpoint) - User wants reports (use ghe-report) - User is not ready to END current phase IMPORTANT: This CLOSES current thread and OPENS next thread. Use ghe-checkpoint if you just want to save progress without changing phases. EXAMPLES: <example> Context: DEV work complete, ready for testing user: "I'm done with DEV, transition to TEST" assistant: "I'll use ghe-transition to close DEV and open TEST thread" </example> <example> Context: All tests pass, ready for review user: "Tests pass, move to REVIEW" assistant: "I'll use ghe-transition to close TEST and open REVIEW thread" </example> <example> Context: Structural issues found in TEST user: "This needs structural changes, demote to DEV" assistant: "I'll use ghe-transition to demote from TEST back to DEV" </example> <example> Context: REVIEW passed, ready to merge user: "REVIEW passed, merge it" assistant: "I'll use ghe-transition to complete the workflow and merge" </example>

Retrieve universal coding patterns from vector database using true two-stage retrieval. Auto-invokes before complex tasks or when user says "--recall". Searches relevant role collections based on task context.

Retrieve universal coding patterns from vector database. Use when starting complex tasks, encountering unfamiliar problems, or user says "--coder-recall" or "--recall" (you decide which scope, may use both). Skip for routine tasks or project-specific questions (use project-memory-recall). This Skill auto-invokes based on task context.

CLAIM a specific GitHub issue to START working on it. Requires an issue number. Performs validation, atomic claim, posts claim comment, and sets up worktree. USE THIS SKILL WHEN: - User says "claim issue #N" or "claim #N" - User says "start working on issue #N" or "take issue #N" - User says "I'll work on #N" or "pick up #N" - User wants to BEGIN work on a specific issue number - User selected an issue from available work and wants to claim it REQUIRED: An issue number must be specified or identifiable from context. DO NOT USE THIS SKILL WHEN: - User just wants to SEE status (use ghe-status) - User is ALREADY working and wants to post update (use ghe-checkpoint) - User wants to CHANGE phases (use ghe-transition) - User wants reports (use ghe-report) - No specific issue number is mentioned EXAMPLES: <example> Context: User wants to start work on a specific issue user: "Claim issue #201" assistant: "I'll use ghe-claim to claim issue #201 with the full protocol" </example> <example> Context: User picks from available work user: "I'll take issue #205" assistant: "I'll use ghe-claim to claim issue #205" </example> <example> Context: User wants to start on specific feature user: "Start working on #312" assistant: "I'll use ghe-claim to claim and set up issue #312" </example>

Expert guide for Model Context Protocol (MCP) integration with Gemini CLI. Covers MCP server configuration (HTTP, SSE, Stdio), connection management, and tool permissions. Use when adding MCP servers to Gemini, configuring transports, troubleshooting MCP connections, or managing tool permissions. Delegates to gemini-cli-docs.

AI-powered audio generation using ElevenLabs API - text-to-speech with lifelike voices, sound effects generation, and music creation from text descriptions. Generate natural-sounding speech in 32 languages, create custom sound effects for games and videos, and compose royalty-free music tracks. Use this skill when the user requests: - Voice generation or text-to-speech conversion - Audio narration for content (videos, audiobooks, podcasts) - Sound effects for games, videos, or applications - Music generation from text descriptions - Multi-speaker dialogue or conversation audio - Voice cloning or custom voice creation - Audio streaming for real-time applications Capabilities: Text-to-speech (32 languages, 100+ voices), sound effects generation, music composition, voice cloning, real-time audio streaming Python SDK: elevenlabs (pip install elevenlabs)

Discipline epistemique stricte pour eliminer les derives cognitives de Claude. Activation SYSTEMATIQUE sur TOUTES les conversations. Impose: (1) Zero invention/supposition, (2) Honnetete absolue, (3) Vision holistique avec relecture du contexte, (4) Questions plutot que suppositions, (5) Remise en contexte perpetuelle avec checkpoints structures, (6) Documentation incrementielle sans doublons, (7) Vigilance workflow/CI-CD/securite, (8) TODO/plan obligatoire avant tout travail, (9) Documentation FR/EN scholar style signee Julien GELEE, (10) Zero credentials/API keys/donnees sensibles en clair. Posture ingenieur generaliste specialise IA.

Autonomous context extraction and structuring for converting unstructured information from multiple sources into AI-interpretable structured formats.

Expert skill for implementing text-to-speech with Kokoro TTS. Covers voice synthesis, audio generation, performance optimization, and secure handling of generated audio for JARVIS voice assistant.