anima_pipeline v1.4.7 – Browser UI for ComfyUI Anima is now available

I've released anima_pipeline — a free, open-source browser UI that automates Danbooru prompt building for ComfyUI Anima workflows.

Type a character name, click buttons to set hair/eyes/outfit/pose, and it sends the prompt straight to ComfyUI. Optional LLM integration (Gemini free tier / LM Studio) auto-generates tags from character names.

→ Full guide (Civitai Article): https://civitai.com/articles/27575 → Download / GitHub: https://github.com/tomotto1296/anima-pipeline → Project page: https://tomotto1296.github.io/anima-pipeline/index_en.html

What's new in v1.4.7:

• Japanese/English UI toggle (auto-detects OS language)

• Log system with API key masking + ZIP export

• Character preset thumbnail gallery

• Bug fixes and stability improvements

Free to use. No account required. Works on Windows, cross-platform via Python.

🗺️ Roadmap

See docs/roadmap.md for the full list.

Coming Soon

• Preset hierarchy (character / scene / camera / quality / LoRA composition)

• Per-preset negative-tag save and auto-switch

• Generation history DB + re-edit support

• Named session saves with multiple slots

• Self-diagnosis UI on failures (error cause hints / missing-node warnings)

• LoRA management improvements (search, favorites, recommended weight, auto-suggestions)

• Batch generation mode (CSV / txt -> sequential generation)

• Generation queue (add / reorder / cancel / rerun)

• Random character preset generation

• Preset sharing (zip Export / Import)

• Prompt diff viewer (vs previous run)

Under Consideration

• First-run tutorial wizard

• Prompt recovery from image metadata (PNG / WebP)

• "Continue with the same character" generation button

• Comparison generation mode (fixed seed for prompt / LoRA / CFG comparisons)

• Workflow reconstruction assistance (semi-automatic mapping)

• LLM-based generated-image scoring and auto-routing

• Auto Danbooru-tag generation from generated images

• Preset sharing via URL code