원클릭으로
conclave
conclave에는 ssube에서 수집한 skills 23개가 있으며, 저장소 수준 직업 범위와 사이트 내 skill 상세 페이지를 제공합니다.
이 저장소의 skills
Launch a Conclave workspace on Runpod. Use when the user wants to start a new GPU pod with the full Conclave environment (Matrix, Planka, Ollama, browser, coding agents, etc).
Connect to the Chromium browser via CDP (Chrome DevTools Protocol) for automation, screenshots, and cookie management. Use when you need to interact with web pages, capture screenshots, or automate browser tasks.
Build, test, and iterate on the Conclave container. Use when making changes to Conclave services, configs, scripts, skills, or extensions and need to rebuild the container image and verify everything works.
Manage tasks and projects in Planka kanban board — create cards, move between columns, track progress. Use when creating tasks, updating card status, checking project boards, managing the kanban workflow, or organizing work items.
Push metrics to Prometheus Pushgateway for Grafana dashboards. Use when recording operational metrics, pushing gauge values to Prometheus, or updating Grafana dashboard data points.
Send messages, read channel history, upload files, manage reactions, and list rooms on Matrix. Use for Matrix communication — posting messages, sharing files, reading conversations, checking reactions, or monitoring rooms.
Send messages, read channel history, upload files, create polls, add reactions, and list guilds on Discord. Use for Discord communication — posting messages, sharing files, running polls, checking conversations, monitoring channels, or listing servers.
Generate images via ComfyUI API with support for Flux and LoRAs, plus image-to-video via Wan 2.2. Use when creating images, generating artwork, running ComfyUI workflows, or producing visual content.
Check the health of all Conclave services. Use when monitoring infrastructure, diagnosing issues, verifying services are running, or during periodic housekeeping checks. Returns structured status for each service.
Semantic search for model descriptions and content using ChromaDB vector database. Use when searching for similar models, finding related content, querying vector collections, or looking up models by description or semantic similarity.
Query ComfyUI server for available models, LoRAs, VAEs, workflows, samplers, and system status. Use when checking what models are loaded, listing available LoRAs on the server, verifying ComfyUI status, or browsing generation capabilities.
Manage ComfyUI jobs — track session jobs, download outputs, check errors, cancel jobs, and monitor queue. Use when checking job status, downloading completed outputs, inspecting failed jobs, or managing the generation queue.
Resize, compress, and format images for platform-specific requirements. Use when preparing images for posting, resizing to meet platform limits, compressing file sizes, or converting between image formats.
Import an Obsidian vault into ChromaDB for semantic search. Parses markdown files, extracts frontmatter, headings, tags, and content sections. Chunks intelligently by heading hierarchy. Use when indexing a knowledge base, importing notes for semantic recall, or syncing an Obsidian vault to a vector database.
Local LLM generation via Ollama — image captioning, prompt generation, text enhancement, and creative writing. Use when captioning images for LoRA training datasets, generating image prompts, enhancing existing prompts, writing model descriptions, or producing creative text without requiring cloud API access.
Interview-driven project planning — structured questioning to produce well-defined specs, then generate Planka cards. Use when planning a new project, breaking down a large initiative, defining requirements through structured questions, or creating a project spec with tasks.
Self-reflection and continuous learning — review context, identify gaps, refine approaches. Use when reflecting on recent work, identifying improvement areas, reviewing alignment, or centering before a creative session.
Create and validate skills for Pi. Use when building a new skill, improving an existing skill's SKILL.md, scaffolding a skill directory, or reviewing skill quality. Covers structure, frontmatter, progressive disclosure, script/reference/asset patterns, and project conventions.
Execute SQL queries against a SQLite database with formatted output. Use when querying data, inspecting schemas, running ad-hoc SQL, or performing pre-defined write operations.
Systematic 4-phase debugging process for any technical issue. Use when encountering bugs, test failures, unexpected behavior, Playwright errors, API failures, ComfyUI issues, session expirations, or infrastructure problems. Finds root cause before attempting fixes.
Capture notes, learnings, and context into ChromaDB for long-term memory. Use when saving something to remember, recording a discovery or decision, storing context for future sessions, or building the knowledge base.
Browse the web — take screenshots, extract text, inspect page elements, and execute JavaScript using Playwright. Use when visiting a webpage, taking a screenshot, extracting text, discovering page structure, or running JS on a site.
Search the web for current information — documentation, troubleshooting, news, platform changes. Use when you need up-to-date information beyond training knowledge, current API docs, error message solutions, platform policy changes, or any factual query where recency matters.