Full index · agent skills
All agent skills — page 2 of 112
Every one of the 11,101 agent skills in the directory, ranked by adoption. Jump to a tool, or browse the agent skills hub.
- Lark SlidesAuthor and theme Lark/Feishu slide decks (e.g. all-hands) using the native presentation XML format.162k installs
- Lark AttendanceQuery your own Feishu/Lark attendance punch records through lark-cli when an agent needs factual check-in history—without asking you for employee IDs.161k installs
- Ai Video GenerationRoute text-to-video, image-to-video, and extend flows across RunComfy models via runcomfy CLI with intent-based model selection.154k installs
- Develop UserscriptsAuthor, debug, and publish Tampermonkey or ScriptCat userscripts with correct metadata, GM APIs, permissions, and ScriptCat-only background or subscription packaging.154k installs
- Ai Image GenerationRoute text-to-image and image-to-image jobs to the right RunComfy model (FLUX, Nano Banana, GPT Image, Seedream, Qwen, Wan) with tuned prompts and `runcomfy run` invokes.154k installs
- Runcomfy CliInstall, authenticate, discover schemas, invoke any RunComfy model endpoint, and poll or download outputs from scripts or the terminal.154k installs
- Face SwapSwap faces or characters in stills or video via RunComfy—routing to Wan Animate, GPT Image Edit, Nano Banana, Flux Kontext, or Kling Motion Control by scenario.153k installs
- Ai Avatar VideoGenerate talking-head, lip-sync, and avatar videos with RunComfy CLI across OmniHuman, Wan, HappyHorse, and Seedance.153k installs
- ImpeccableRedesign or polish dashboards, landings, forms, and components with UX critique, a11y, motion, and live browser iteration.152k installs
- Video InpaintingLet agents remove objects, clean watermarks, or patch masked regions across video frames using the RunComfy CLI with the right Wan, Lucy Edit, or Seedream route.151k installs
- Image InpaintingRemove objects or watermarks from still images using a mask,Fill or replace a masked region with Z-Image Turbo or edit models,Run inpainting jobs through the runcomfy CLI from an agent workflow151k installs
- Controlnet PoseTransfer motion from a reference video onto a target character (Kling motion control),Generate images conditioned on OpenPose, DWPose, depth, or canny references,Pick video vs still and photoreal vs s151k installs
- LipsyncAnimate a portrait still plus audio into a speaking avatar (OmniHuman),Re-sync mouth movement on existing video to a new audio track (Sync Labs, Kling, Creatify),Generate video with synced speech from151k installs
- Video ExtendExtend short Veo clips into longer sequences or chained shots via RunComfy CLI extend-video endpoints.151k installs
- Image OutpaintingUncrop stills, change aspect ratio, and expand canvas while keeping the original subject intact via RunComfy edit routes.150k installs
- Video OutpaintingSpatially widen or reframe video (e.g. 9:16 to 16:9) while keeping central action consistent.150k installs
- RelightRelight product shots or portraits via RunComfy CLI without reshooting—studio, golden hour, rim light, and color-temperature changes.150k installs
- Elevenlabs Music GenerationGenerate vocal songs, instrumentals, jingles, podcast intros, or game loops from prompts via ElevenLabs Music on RunComfy.150k installs
- Lark OkrView and manage Lark/Feishu OKR cycles, objectives, key results, alignment, and progress records from the CLI146k installs
- Ace StepGenerate, inpaint, or extend stereo music tracks with StepFun ACE Step through the RunComfy CLI when you need cheap, tag-driven audio for demos, apps, or ads.142k installs
- Ai MusicLet the agent pick RunComfy’s ElevenLabs or ACE Step music stack from plain-language intent—vocals, cheap beds, multilingual lyrics, or inpaint/outpaint edits—and run the right `runcomfy` command.142k installs
- PrototypeBuild a small terminal prototype to manually drive state transitions and verify business logic and edge cases before implementing production code.140k installs
- Caveman CommitGenerate ultra-terse Conventional Commits messages from staged changes, with a short body only when the rationale is not obvious from the subject.140k installs
- Paper Context ResolverResolve narrow paper-backed gaps (splits, preprocessing, eval protocol, checkpoints, runtime) when README and repo files are insufficient for faithful ML paper reproduction.140k installs
- Repo Intake And PlanScan an ML research repository, extract README commands, and recommend the smallest trustworthy inference or evaluation reproduction target.140k installs
- Env And Assets BootstrapBootstrap conda-first environments plus checkpoints, datasets, and cache paths before a documented reproduction run.140k installs
- Minimal Run And AuditRun documented inference, evaluation, or smoke commands and normalize evidence into standardized repro_outputs with patch notes.140k installs
- Caveman ReviewProduce terse, line-scoped PR review comments with optional severity tags instead of verbose feedback.139k installs
- PptxCreate, read, and edit PowerPoint (.pptx) decks from the agent when pitch materials, status reports, or launch slides must land as real Office files—not screenshots.138k installs
- Caveman CompressShrink CLAUDE.md and memory files into caveman format to cut input tokens while keeping backups.137k installs
- Vercel React Native SkillsApply prioritized React Native performance rules (lists, rendering, stable refs) when agents generate or refactor mobile UI code.137k installs
- HandoffEnd a long agent session by writing a compact handoff doc (with suggested skills) so the next agent continues without re-reading the full thread.136k installs
- Using SuperpowersInstall this as the default ritual so your coding agent always loads the right SKILL.md before planning, coding, or answering—instead of guessing from chat alone.134k installs
- Systematic DebuggingInstall this when you want your coding agent to investigate root causes before guessing fixes on failures, bugs, or flaky behavior.134k installs
- Writing PlansTurn a written spec into a file-level, bite-sized implementation plan your coding agent can execute without guessing about your repo.133k installs
- Caveman HelpOne-shot cheat sheet of Caveman slash commands, modes, sibling skills, and how to deactivate or set default mode—without changing active mode.133k installs
- Just ScrapeInstall when you need search, scrape, crawl, structured JSON extraction, or change monitoring from URLs without wiring ScrapeGraph into app code first.133k installs
- Seo AuditRun a structured technical and on-page SEO audit on your marketing site when organic traffic feels underwhelming and you need a prioritized fix list.131k installs
- PdfInvoke whenever PDFs show up—merge contracts, extract tables, fill forms, OCR scans, or ship a new PDF deliverable from agent workflows.130k installs
- Lark MarkdownCreate, fetch, patch, diff, and overwrite Markdown files stored in Lark/Feishu Drive via lark-cli instead of using docx or generic drive flows.125k installs
- Airunway Aks SetupWalk from an existing AKS cluster through AI Runway controller install, GPU assessment, provider setup, and first model deployment with cost and prerequisite handoffs.124k installs
- Design Taste FrontendShip polished React/Next.js UIs with explicit design-variance, motion, and density knobs; enforce dependency checks, RSC defaults, and metric-driven layout instead of generic LLM styling.124k installs
- CopywritingDraft or improve conversion-focused copy for homepages, landing pages, pricing, features, and about pages with audience and single-primary-CTA framing.121k installs
- Requesting Code ReviewDispatch a subagent with a senior code-reviewer prompt, your plan, and a git SHA range so finished work is checked before you merge or stack more changes.119k installs
- DocxCreate, read, edit, or repack Word .docx deliverables (reports, memos, letters, TOC, tracked changes, images) via pandoc, docx-js, or XML unpack/repack—not PDFs or spreadsheets.118k installs
- Extract Design SystemScrape design primitives from a public reference site into normalized JSON and starter CSS token files for your app.118k installs
- Test Driven DevelopmentMake your coding agent follow strict test-first red-green-refactor before any feature, bugfix, or refactor production code.118k installs
- Seedance V2cinematic-short-form-video114k installs
- Image To Videostill-to-motion-animation113k installs
- Video EditRoute video edit intents (restyle, background swap, motion control) to the right RunComfy model via CLI.113k installs
- Ai Image GenerationText-to-image and image-to-image across FLUX, Nano Banana, GPT Image, Seedream, Qwen,Choose models for typography, photoreal portraits, or fast iteration,Restyle or edit images with multi-reference br113k installs
- Codex PetCreate Codex-compatible pet spritesheets and pet.json from one image and install under CODEX_HOME pets.113k installs
- Runcomfy CliInstall, authenticate, and invoke RunComfy model endpoints from scripts or agent shells with poll, stream, and JSON output.113k installs
- Image OutpaintingExtend still image canvas, uncrop, and change aspect ratio while preserving the original scene via Nano Banana 2, GPT Image 2, FLUX Kontext Pro, or brand-locked edit endpoints.113k installs
- Controlnet PosePose-, depth-, and motion-conditioned image or video generation through RunComfy CLI routing.113k installs
- Face SwapSwap faces or characters in stills or video by choosing the right RunComfy model (Wan Animate, GPT Image 2 Edit, Nano Banana, Flux Kontext, Kling Motion Control) from user intent.113k installs
- Nano Banana EditEdit images with Google Nano Banana 2 on RunComfy via the local CLI, with routing guidance to sibling edit models when another endpoint fits better.113k installs
- Ai Avatar VideoGenerate talking-head, lip-sync, and avatar videos from audio or portraits via RunComfy CLI across OmniHuman, Wan 2-7, HappyHorse, and Seedance v2.113k installs
- Gpt Image 2typography-and-brand-imagery113k installs
- Face SwapSubstitute one identity for another in stills or video by routing RunComfy CLI jobs to Wan animate, GPT Image 2, Nano Banana, Flux Kontext, or Kling motion models.113k installs
- Video OutpaintingExtend video spatial canvas and convert aspect ratios (e.g. 9:16 to 16:9) while keeping central action, via Wan 2-7 edit-video or ComfyUI outpaint workflows for hero seams.113k installs
- Elevenlabs Music GenerationCreate songs, instrumentals, jingles, and podcast/game audio from text via ElevenLabs Music on RunComfy with section tags and duration control.113k installs
- Happyhorse 1 0Generate text-to-video with HappyHorse 1.0 on RunComfy—1080p, in-pass synced audio, multi-shot consistency—using documented duration, aspect ratio, and resolution options.113k installs
- Image InpaintingRemove objects, watermarks, or blemishes with a binary mask,Fill or replace a masked region via Z-Image Turbo Inpainting,Describe a region in prose when no mask using Nano Banana, GPT Image, or FLUX K113k installs
- Runcomfy CliInstall and authenticate the RunComfy CLI,Discover model schemas and invoke image/video/edit endpoints,Poll job status and download outputs from scripts113k installs
- Gpt Image EditEdit existing images with GPT Image 2 on RunComfy CLI using preservation-aware prompts and multi-reference inputs113k installs
- Video ExtendExtend or continue existing Veo 3-1 clips via RunComfy CLI, chain narrative shots from a seed video, and produce longer clips with consistent motion and subject identity.113k installs
- Image InpaintingMask-driven inpainting and region edits on RunComfy—object or watermark removal, blemish cleanup, fill and replace masked areas—with Z-Image Turbo Inpainting or prose-driven edit models when no mask e113k installs
- RelightRelight product or portrait stills via RunComfy (Qwen relight LoRA or fallback edit models) without reshooting.113k installs
- Nano Banana 2Generate images with Google Nano Banana 2 on RunComfy for rapid iteration, in-image typography, social thumbnails, and drafts; includes resolution pricing, safety tolerance, and routing guidance to Pr113k installs
- Gpt Image 2Generate or edit marketing and product visuals with OpenAI GPT Image 2 through RunComfy’s CLI without wiring your own OpenAI API key.113k installs
- LipsyncMatch mouth movement to an audio track on portrait stills or existing video via RunComfy lip-sync endpoints113k installs
- Controlnet PoseRoute pose-, skeleton-, depth-, or motion-conditioned image and video jobs to the right RunComfy endpoints (Kling motion control, Wan Animate, Z-Image Turbo ControlNet) from natural-language requests.113k installs
- Video InpaintingRemove objects or watermarks across video frames,Replace masked regions with motion-consistent fills,Pick Wan, Lucy Edit, or Seedream routes by edit style113k installs
- Flux 2 KleinFast image generation with Flux 2 Klein on RunComfy, including 9B/4B choice, step-count tuning, multi-reference styling, and fallbacks to Flux 2 Pro or Seedream.113k installs
- Elevenlabs Music GenerationCompose full vocal songs or instrumental beds from text with ElevenLabs Music on RunComfy for podcasts, games, ads, and product videos.113k installs
- Ai Video GenerationGenerate or extend AI video clips from text or images by auto-selecting HappyHorse, Wan, Seedance, Kling, Veo, Hailuo, and related RunComfy models113k installs
- Ai MusicGenerate vocal songs, instrumentals, and jingles with RunComfy CLI,Pick ElevenLabs vs ACE Step by budget and quality needs,Inpaint or outpaint a time range in an existing track113k installs
- Wan 2 7Generate text-to-video with Wan 2.7 on RunComfy, including multi-reference motion and audio-driven lip-sync when audio_url is set.113k installs
- Flux 2 KleinGenerate fast Flux 2 Klein images on RunComfy with distilled-model prompting, step-count strategy, 9B vs 4B choice, and fallbacks to Flux 2 Pro, Seedream 5, or GPT Image 2.113k installs
- LipsyncSync a portrait still plus audio into talking avatar video,Re-mouth an existing clip to new voiceover audio,Generate speech-driven video from script via routed Kling or Creatify endpoints113k installs
- Ace StepGenerate or edit stereo music from tags and lyrics through RunComfy’s ACE Step APIs without wiring your own StepFun hosting.113k installs
- RelightRelight product shots, portraits, or any still via RunComfy CLI with Qwen relight LoRA or fallback edit models.113k installs
- Video ExtendLengthen an existing Google Veo 3.1 clip or chain the next narrative beat via RunComfy’s extend-video CLI without hand-rolling video API jobs.113k installs
- Ai Video GenerationGenerate or extend AI video clips from text or images by auto-selecting the best RunComfy video model for quality, audio, motion, or speed.113k installs
- Nano Banana 2Generate flash-tier Google Nano Banana 2 images on RunComfy with documented prompting, typography, and routing guidance instead of naive one-line prompts.113k installs
- Image EditRoute image-edit requests to the best RunComfy edit model (batch identity edits, multilingual text-in-image, high-fidelity local edit, or mask inpaint) and execute via CLI with bundled per-model promp113k installs
- Ai Avatar VideoCreate talking-head, lip-sync, UGC, and virtual presenter videos by routing to the right RunComfy avatar model and running documented runcomfy CLI invokes113k installs
- Flux KontextRun precise single-reference local image edits with Flux 1 Kontext Pro on RunComfy, including when to switch to Nano Banana, GPT Image 2, or Flux 2 Klein instead.113k installs
- Kling 3 0Generate cinematic multi-shot video with Kling 3.0 on RunComfy across Standard, Pro, and 4K text-to-video and image-to-video tiers via the CLI.113k installs
- Ai Image GenerationRoute text-to-image and image-to-image jobs to the right RunComfy model (FLUX, Nano Banana, GPT Image 2, Seedream, Qwen, Wan) with documented prompts and a single CLI invoke.113k installs
- Image To VideoAnimate a still image on RunComfy by routing to HappyHorse, Wan 2.7, or Seedance 2.0 Pro with the right prompts and audio options via the CLI.113k installs
- Image OutpaintingExtend still images beyond their original frame on RunComfy—uncrop, change aspect ratio, and fill missing canvas—without hand-picking which edit model to call.113k installs
- Image EditRoute everyday image edits—background swap, object removal, headline rewrite, batch i2i—through the best RunComfy edit model and its prompting patterns via one skill.113k installs
- Happyhorse 1 0Generate native 1080p text-to-video with synchronized audio via HappyHorse 1.0 on the RunComfy CLI, with guidance on duration, aspect ratio, and when to use other video models.113k installs
- Seedance V2Generate 4–15s cinematic short-form video with Seedance 2.0 Pro on RunComfy using multi-modal image, video, and audio references and native lip-sync113k installs
- Flux KontextRun precise single-reference local image edits with Flux 1 Kontext Pro via RunComfy CLI, using model-specific prompting and routing guidance when other edit models fit better.113k installs
- Gpt Image EditEdit images with OpenAI GPT Image 2 on RunComfy via CLI, using preservation and layout prompting patterns and routing to alternate editors when appropriate113k installs
- Kling 3 0Generate multi-shot cinematic Kling 3.0 video with native audio from text or a reference image through the RunComfy CLI.113k installs
- Video EditRoute a solo builder’s video-edit intent to the right RunComfy model and run edits via the local CLI without trial-and-error across Wan, Kling, and Lucy.113k installs