Three states hold ComfyUI together: how SUCCESS, FAILURE, PENDING keep graph execution correct
Mahmoud_Zalt · x · 2026-09-25
AI solutions architect Mahmoud Zalt published a deep dive into ComfyUI's execution engine and how it keeps graph-shaped prompts correct when nodes wait, fail, or spawn more work:
- Core design: ComfyUI executes prompts as dependency graphs, not linear scripts; execution.py turns graphs into scheduled node work, and its most consequential choice is small — every execution attempt reports SUCCESS, FAILURE, or PENDING
- From validation to runnable work: validateprompt recursively checks registered classes, link shapes, types, required values, custom validation, cycles and loop boundaries before PromptExecutor stages nodes via DynamicPrompt, cache sets, and ExecutionList
- PENDING as a correctness state: the article traces how readiness is established and which boundaries must uphold the three-state contract as the system grows
Useful for workflow-engine design or ComfyUI node developers.
More from coding & agent
- Pose Blueprint: A Browser-Based 3D Pose Editor for ComfyUI and ControlNet — OkConfusion6667 · 2026-09-25
- DigitalOcean hosts classifier model Jev as a first-class agent tool, not an endpoint — hardimanjames · 2026-09-25
- The wiki is an anti-pattern: docs must live in the repo for coding agents to maintain — TechPreacher · 2026-09-25
- Caching policy lookups per inode cuts eBPF security agent CPU cost ~90% — JeremyCMorgan · 2026-09-25
- jev-test-impact uses an LLM to pick which tests to run from a Git diff — alchemist-301 · 2026-09-25
- Astronomer builds his dream learning tool in ~5 hours with Opus 5.5, now live — niloofar_mire · 2026-09-25