Total 50,706 skills, AI & Machine Learning has 8496 skills
Showing 12 of 8496 skills
Literature Scout — Responsible for multi-source literature retrieval, screening, and classification, and constructing literature matrices. Activated when assigned by research supervisors to collect literature. Conduct systematic literature retrieval using tools such as Exa, ArXiv API, Semantic Scholar, etc.
Async music / audio-track generation via Venice. Covers the /audio/quote + /audio/queue + /audio/retrieve + /audio/complete lifecycle, lyrics vs instrumental, voice selection, duration, language, speed, model capability probing, and webhook-free polling.
This skill should be used when the user asks to "fix my skill" or "audit this skill". Make sure to use this skill whenever the user mentions skill quality, structural issues, broken skills, or skill diagnostics — even if they don't explicitly say "repair-skill". Not for adding features or improving effectiveness — use improve-skill. Not for agents — use repair-agent.
Wide before deep. Fans out N parallel divergent thoughts under structurally different cognitive frames (regulator, biology, speedrunner, 10 year old, $0 budget), then scores, clusters, prunes traps, and deepens only the top survivors. The isolated parallel branches and the separated generator/critic phases are load-bearing. Do not collapse them into a single linear thought. Use when the user asks to brainstorm, ideate, generate options, design an architecture, name something, pick between approaches, plan a refactor, design an API or SDK surface, generate hypothesis classes for a fuzzy bug, or any prompt of the shape "give me a few ways to". Also use when the obvious answer feels obvious and wrong, or when the user explicitly invokes /adhd or asks for "ADHD mode".
Builds and debugs Letta Code channels, including first-party channel adapters and dynamic user channel plugins under ~/.letta/channels. Use when adding Telegram, WhatsApp, Bluesky, Slack, Discord, or custom channel support; testing channel routing, pairing, MessageChannel, runtime dependencies, or channel plugin manifests.
Build high-quality Agent Skills for Claude following official Anthropic best practices. Covers SKILL.md structure, frontmatter, description writing, progressive disclosure, testing, patterns, troubleshooting, and distribution across all surfaces (Claude.ai, Claude Code, API, Agent SDK). Use when creating new skills, reviewing skill quality, debugging skill triggering, structuring skill directories, writing skill descriptions, or improving existing skills. Triggers on "build a skill", "create a skill", "skill structure", "SKILL.md", "skill best practices", "skill not triggering", "skill quality".
Control image generation requests before execution. Use this when the user wants text-to-image, image edit, reference-image generation, product image, persona image, banner, thumbnail, storyboard image, or image batch variants and the skill must identify inputs, classify the task, choose model/reference rules, then hand off to image-batch-runner.
Control video generation requests before execution. Use this when the user asks for a simple clip, storyboard video, UGC video, podcast clip, reference video, talking-head, image-to-video, text-to-video, or research-handoff video and the skill must classify the request before handing it to video-request-architect and a runner such as seedance-submitter or video-batch-runner.
Interactive QA session where users report bugs or issues through conversation, and the agent creates GitHub issues. Explore the codebase in the background to obtain context and domain language. Use when user wants to report bugs, do QA, file issues conversationally, or mentions "QA session".
Automated AI content pipeline from research to video generation using Claude/OpenAI and Remotion
Two-host podcast video for any URL or free-form topic — 1 minute, 4 acts × ~15s, native multi-shot dialogue, optional voice cloning for Host A. Use when the user asks to "make a podcast", "podcast about [thing]", "podcast review of [url]", "two-host explainer", "interview-style clip", "two people talking on camera", "I/me and X talk about Y", or "interview with [persona] about [topic]". Native audio is the deliverable; captions are skipped by default because podcast dialogue mistranscribes domain terms.
Filesystem RAG benchmarks: corpus/, train.json, evaluate_rag.py (RAGAS quality). Not for prod monitoring, latency/throughput benchmarking (use rag-perf), or evals outside this repo layout.