Headless agentic image gen with n8n + llama-swap: where the pipeline hits its limits
WhiskeyBatman · reddit · 2026-10-20
WhiskeyBatman details his Discord bot that manages game servers in-character: n8n on a separate server orchestrates calls to a headless inference machine running llama-swap, which starts/stops llama or vllm as needed plus Stable Diffusion. Image calls use flux.2 klein 9b with reference images — an LLM writes descriptions of changes relative to the reference, so prompts stay minimal.
Upgrading is proving hard: the SD stack can't do KREA2 identity edit with reference images as latents (always falls back to vision, even with LoRA), and Comfy UI over HTTP/MCP works but feels too heavy and too "local first." He's asking what other agentic image-gen pipelines people are running.
More from coding & agent
- Index Network open-sources Social Discovery Protocol: agents negotiate fit on your behalf — carsonfarmer · 2026-10-23
- Installed a formal-verification skill: it silently installed Java, MacBook runs 10 degrees hotter — zealcaiden · 2026-10-23
- OpenAI's Decisions API hits GA: near real-time routing up to 10x faster than GPT-6 Luna — OpenAIDevs · 2026-10-23
- Ruff author warns: upgrade Node/Bun now — --use-system-ca still accepts the eMudhra cert — charliermarsh · 2026-10-23
- Proof of Agent: paste one prompt and your AI mines, grades, and self-governs — adamamcbride · 2026-10-23
- DeepLearning.AI weekly: GPT-6.1 hits 97.9% byte parity on COBOL migration — DeepLearningAI · 2026-10-23