<h1 id="module-flow">🛠️ Module Flow</h1>A desktop application for managing generative pipelines. It lets you build modular prompt chains, control generation parameters, run tasks, and track results - all in one interface.Module Flow is a standalone application that connects to external engines like ComfyUI via API. It's designed for people who want full control over prompts, parameters, and outputs.The goal is to make generation reproducible, configurable, and understandable, especially in workflows with multiple variations, iterations, and parameter sweeps.<h2 id="core-features:">🧩 Core features:</h2>- Prompt chains built from modular blocks with text, numbers, ranges; supports cycles, randomness, prefixes/postfixes.- Parameter linking: any prompt value can drive generation settings like seed, resolution, CFG, steps, etc.- Task management: queue system, status tracking, previews, auto-saving all configs and outputs.- API integration: currently supports ComfyUI; architecture is open to extension.<h2 id="use-cases:">📌 Use cases:</h2>- Generative content factories: define prompt logic + auto parameter variation = dozens of outputs.- Model testing: run controlled experiments across different seeds, resolutions, CFG, etc. with no manual copying.- Preset creation and iteration: save, reuse, and switch between prompt/config snapshots.- Asset production for games/media: swap topics, styles, or objects at scale with consistent results.<h2 id="who-it's-for:">🎯 Who it's for:</h2>- AI tinkerers and researchers - run large-scale prompt experiments with full traceability and parameter logic.- Artists and visual creators - structure ideas, automate variation, and get repeatable results with aesthetic control.- Studios and production teams - manage content pipelines in one place: templates, presets, task history, and reusable workflows.<h2 id="download:">📦 Download:</h2>Module Flow is available here on CivitAI and on <a target="_blank" rel="ugc" href="https://github.com/Pecha85/ModuleFlow">GitHub</a> Linux and macOS builds are in development and will be released soon<h2 id="feedback:">💬 Feedback:</h2>Module Flow <a rel="ugc" href="https://discord.gg/2czd6aSah3">Discord </a>server.Join to:<ul><li>Discuss features, updates, and roadmap direction</li><li>Get help with setup, bugs, and advanced configurations</li><li>Share prompt chains and workflow presets</li><li>Showcase results and techniques</li><li>Connect with artists, tech specialists, and studios using generative pipelines</li></ul>This is the space to streamline, systemize, and scale your AI-powered production.Feedback, feature suggestions, and bug reports are very welcome. The tool is in early development, and input helps prioritize the right things. Feel free to open <a target="_blank" rel="ugc" href="https://github.com/Pecha85/ModuleFlow/issues">GitHub issues</a> or comment here.⚠️ Please note: this is an early-stage tool. Bugs and missing features are expected - reports are appreciated.- Support via <a target="_blank" rel="ugc" href="https://www.patreon.com/PechaSD">Patreon</a> is appreciated if you find this tool useful (Boosty coming soon).

Module Flow - Desktop application for managing generative pipelines.

Fixed- Text display adjustments in UI- UI module delamination when task preview is enabled and the chain is hidden- The ability to delete tasks during generation- Rename "prompt_group" → "module_presets". Remove "exaples" folder and restart for update- Task Resource name fixes- "Wrong JSON" error during config openingAdded- Drag&amp;Drop JSON system. Automatically opened modules, chain presses, generative configs and tasks when D&amp;D files to the interface- Warnings when starting a task without configs and links- Updates in the "info" section- Added check for the existence of the task folder before pressing the button- JSON protection system for config loading- Warnings with JSON type when open config- Documentation updates

Files with SHA256: 58e76c0e72ec1296ba285c52d21173f29e9f6dc7712e9ae510770ac0fcc6d8bb