hiteshbandhu/skills-i-use · GitHub
Builds and ships generative image/video products—FLUX, ComfyUI workflows, Veo, fal platforms, Luma-scale launches, and API-first creator UX. Use when working on diffusion pipelines, Comfy graphs, video models, generative media infra, or viral consumer media launches.
프로젝트 폴더에서 아래 명령어를 실행하고, 설치할 에이전트를 선택하세요.
npx skills add hiteshbandhu/skills-i-use --skill build-generative-media설치 명령을 직접 실행해야 적용됩니다. 지원 에이전트와 필요한 권한·라이선스는 제작자의 안내를 확인하세요.
README.md# Build generative media Playbooks for FLUX, ComfyUI, Luma scaling, fal platforms, Veo/GenMedia, and creator-facing gen-AI APIs. ## Usage ``` @build-generative-media ComfyUI workflow CI plan Luma-style scale checklist before launch ``` ## Install ```bash cp -r skills/build-generative-media ~/.claude/skills/ ``` Source: `playlists/generative-media-ai-engineer/` in [ingest-into-skills](https://github.com/hiteshbandhu/ingest-into-skills).
SKILL.md--- name: build-generative-media description: > Builds and ships generative image/video products—FLUX, ComfyUI workflows, Veo, fal platforms, Luma-scale launches, and API-first creator UX. Use when working on diffusion pipelines, Comfy graphs, video models, generative media infra, or viral consumer media launches. --- # Build generative media Action playbook from nine AI Engineer generative media talks. **Do not summarize** — pick a workflow theme. **Supporting files:** [workflows.md](workflows.md) · [source-index.md](source-index.md) Optional: `./skill-outputs/build-generative-media/` --- ## Step 0 — Pick workflow ``` What is the user trying to do? ├─ Choose image model (FLUX, open research) → image-models ├─ ComfyUI reproducible node pipelines → comfy-workflows ├─ Viral consumer launch / GPU scale (Luma) → scale-launch ├─ Creator API UX (Replicate-style) → api-ux ├─ Multi-model platform routing (fal) → media-platform ├─ Google Veo / GenMedia enterprise chains → google-video ├─ Diffusion training/serving at scale (DeepMind) → training-scale └─ Exploration mindset for new modalities → curiosity-lab ``` --- ## Install ```bash cp -r skills/build-generative-media ~/.cursor/skills/ cp -r skills/build-generative-media ~/.codex/skills/ ``` Source: `playlists/generative-media-ai-engineer/`. --- ## Cross-cutting rules | Rule | Source | |------|--------| | Embed workflow JSON in outputs for reproducibility | [src-002 @ 0:01:58] | | Rehearse 10x scale before viral marketing | [src-003 @ 0:01:37] | | Video compute ≠ image—budget tokens separately | [src-007 @ 0:04:52] | | Enterprise gen media needs brand/safety defaults | [src-006 @ 0:15:04] | | Demoable, legible APIs beat opaque pipelines | [src-004 @ 0:01:15] | --- ## Output Name theme; artifacts under `./skill-outputs/build-generative-media/` when requested.
source-index.md# Source index — Build generative media Notes: `playlists/generative-media-ai-engineer/learnings/`. | ID | Talk | Learning file | |----|------|----------------| | src-001 | FLUX & visual AI (Stephen Batifol, BFL) | `learnings/flux-open-research-and-the-future-of-visual-ai-stephen-batifol-black-forest-labs.md` | | src-002 | ComfyUI workshop (ComfyAnonymous) | `learnings/comfyui-full-workshop-first-workshop-from-comfyanonymous-himself.md` | | src-003 | Dream Machine scale (Keegan McCallum, Luma) | `learnings/dream-machine-scaling-to-1m-users-in-4-days-keegan-mccallum-luma-ai.md` | | src-004 | Design like Karpathy is watching (Zeke Sikelianos) | `learnings/design-like-karpathy-is-watching-zeke-sikelianos-replicate.md` | | src-005 | State of generative media (Gorkem Yurtseven, fal) | `learnings/the-state-of-generative-media-gorkem-yurtseven-fal.md` | | src-006 | Bananas / GenMedia (Guillaume Vernade) | `learnings/let-s-go-bananas-with-genmedia-guillaume-vernade-google-deepmind.md` | | src-007 | Generative image/video at scale (Sander Dieleman) | `learnings/building-generative-image-video-models-at-scale-sander-dieleman-google-deepmind.md` | | src-008 | Veo 3 for developers (Paige Bailey) | `learnings/veo-3-for-developers-paige-bailey-google-deepmind.md` | | src-009 | On Curiosity (Sharif Shameem, Lexica) | `learnings/on-curiosity-sharif-shameem-lexica.md` | Playlist: https://www.youtube.com/playlist?list=PLcfpQ4tk2k0VjKRy3q6ZxeOtkbZlmFDLg
workflows.md# Workflows — Build generative media --- ## Workflow: image-models **Use when:** selecting FLUX or similar image backbones. ### Steps 1. Match FLUX variant (base, Context) to gen vs edit use case [src-001 @ 0:02:13] 2. Clarify enterprise licensing vs open weights [src-001 @ 0:01:05] 3. Benchmark on domain imagery before commit [src-001 @ 0:05:49] ### Agent deliverables - [ ] Model selection memo with eval samples --- ## Workflow: comfy-workflows **Use when:** productionizing node-graph media pipelines. ### Steps 1. Build modular graphs (ControlNet, img2img) [src-002 @ 0:01:51] 2. Embed workflow metadata in outputs [src-002 @ 0:01:58] 3. Pin workflow versions for CI reproduction [src-002 @ 0:02:27] ### Agent deliverables - [ ] Versioned Comfy workflow JSON + runbook --- ## Workflow: scale-launch **Use when:** consumer launch may go viral. ### Steps 1. Run staged 10x capacity drills with monitoring [src-003 @ 0:01:37] 2. Plan queueing and repeated spike handling [src-003 @ 0:01:57] 3. Harden GPU path pre-marketing [src-003 @ 0:03:08] ### Agent deliverables - [ ] Launch scale runbook (capacity steps, alerts, rollback) ### Stop if - No load test before social/PR push [src-003 pitfalls] --- ## Workflow: api-ux **Use when:** creator-facing API or product surface. ### Steps 1. Optimize for fast visual preview feedback [src-004 @ 0:01:15] 2. Label model capabilities clearly in catalog [src-004 @ 0:02:00] 3. Demo flows as if expert reviewers will judge [src-004 @ 0:14:18] ### Agent deliverables - [ ] API UX spec with preview latency targets --- ## Workflow: media-platform **Use when:** routing multiple models via a platform (fal-class). ### Steps 1. Route workloads across model catalog [src-005 @ 0:00:32] 2. Tune stacks per vertical (ads, fashion, gaming) [src-005 @ 0:05:27] 3. Track new open weights (FLUX, etc.) on platform [src-005 @ 0:04:37] ### Agent deliverables - [ ] Multi-model routing matrix + SLA tiers --- ## Workflow: google-video **Use when:** Veo 3 or GenMedia enterprise chains. ### Steps 1. Integrate Veo via generative media APIs [src-008 @ 0:00:49] 2. Check watermark/rights policies [src-008 @ 0:08:36] 3. Chain image→video→upscale for campaigns [src-006 @ 0:10:57] 4. Enable enterprise safety defaults on prompts [src-006 @ 0:15:04] ### Agent deliverables - [ ] GenMedia pipeline diagram + compliance checklist --- ## Workflow: training-scale **Use when:** training/serving large diffusion video. ### Steps 1. Model video token cost vs images [src-007 @ 0:04:52] 2. Use coarse-to-fine schedules for video [src-007 @ 0:08:55] 3. Align codecs/resolution with infra budget [src-007 @ 0:05:11] ### Agent deliverables - [ ] Training cost estimator (resolution × duration × FPS) --- ## Workflow: curiosity-lab **Use when:** exploring new gen-AI product ideas. ### Steps 1. Ship small demos per model capability [src-009 @ 0:07:03] 2. Follow non-obvious use cases via experimentation [src-009 @ 0:02:29] 3. Avoid only building obvious wrappers [src-009 @ 0:09:26] ### Agent deliverables - [ ] Demo backlog with learnings per experiment