ComfyUI CoolB Nodes

ComfyUI CoolB Nodes ⚡️

A simple pack of utility nodes for ComfyUI featuring saving/loading speakers with transcripts (useful alongside TTS packs), execution flow control (muting/bypassing nodes and groups on trigger to actively change the workflow state from run to run), some useful image, audio, video and text file utilities (including save/load text file) and data type conversion.

This pack was created primarily to support my other project — a fully Autonomous Telegram Bot inside a single ComfyUI workflow.

---

🧩 List of Nodes

<img width="2083" height="972" alt="image" src="https://github.com/user-attachments/assets/e6f3b9e1-d1a6-48d2-b687-f11c6664d3cd" />

🎙️ Speaker Management

Designed for TTS (Text-to-Speech) workflows, voice cloning, and audio archiving. Automatically uses PyAV or torchaudio fallback.

• CB Save Speaker: Saves an audio tensor/file alongside its provided text transcription (speech script) into a structured format inside models/SPEAKERS (customizable via config.json). Supports .mp3 and .wav.
• CB Load Speaker: Lists and loads saved speakers dynamically from your storage, outputting both the audio waveform and its reference text.

🔊 Audio processing

• CB Audio Info: A simple node that provides information about the input audio (duration, sample rate, total samples, number of channels).
• CB Audio Processor: Processes an audio track with volume, fades, and silence padding. Can be used alongside native "Merge Audio" nodes to combine multiple tracks with precise control.
• CB Roll and Trim Audio Samples: Rolls audio by a time offset first, then trims from start/end. No resampling is performed, does a cyclic sample shift and sample-accurate trim. Returns trimmed audio, removed start audio, and removed end audio. Can be used to trim and split audio track at the right position or when dealing with looping videos.

🎬 Image/video processing

• CB Roll and Trim Video Frames: Does the same as the other node for audio, but for video frames. Rolls the clip by an integer offset first, then trims frames from the start/end. Returns trimmed video, removed start frames, and removed end frames. Very useful for working with looping videos.
• CB Nearest Resolution: Calculates the nearest resolution for an input image/video that matches chosen megapixels and divisible_by, while keeping the aspect ratio closest to the original. Useful for quick calculations of width/height. The output can be connected to the native "Resize Image/Mask" node.
• CB Color Correct: Allows to tweak the color of the image/video by changing the following parameters: temperature, hue, brightness, contrast, saturation and gamma. An improved version of the node from from art-venture pack; was optimized and made much faster for processing batches.

⚙️ Automation & Workflow Control

• CB Set Mute/Bypass State: Allows you to dynamically Mute, Bypass, or Activate one or multiple nodes specified by their IDs using any passthrough signal (applies to the next prompt run). An improved version of the node from ComfyUI-Impact-Pack.
• CB Set Group Mute/Bypass State: Same as the other node but for entire groups. Changes execution states (Mute/Bypass/Active) for all nodes within specified groups at once based on a group name search string (applies to the next prompt run). Node detection adapted from rgthree-comfy group muters/bypassers.

📝 Text & String Utilities

• CB Textbox: A simple text container that displays text, prints to screen and automatically overwrites its contents and converts any incoming data stream to a string. An improved version of the node from ComfyUI-Chibi-Nodes.
• CB String Selector: Splits a multiline or delimited string (by newlines, commas, or spaces) and returns a specific substring based on an index (supports modulo wrapping).
• CB String List Match Index: Compares an input string against a list of strings that are devided by a chosen delimiter. Supports custom matching modes (Equal, Contains, Starts with, Ends with) and returns a 1-based index (0 if not matched) and a boolean match state.

💾 Text Save/Load

• CB Read Text File: Loads raw text content from any specified file path relative to your ComfyUI output folder.
• CB Save Text File: Saves text strings into files with three operation modes: overwrite (replace), append (add to end), or increment (automatically generates numbered filenames like config_00001.txt). Also limited to ComfyUI output folder for safety reasons.

🔄 Type Conversion

• CB Data type Converter: Accepts any input connection type and safely converts it to a standard primitive: STRING, FLOAT, INT, or BOOLEAN.

---

🛠️ Installation and requirements

Install Via ComfyUI Manager (clones the repo and installs requirements.txt)

1. Open ComfyUI and click on the Manager button.
2. Click Custom Nodes Manager.
3. Search for ComfyUI CoolB Nodes and click Install.
4. Restart ComfyUI.

Requirements: torch, av, numpy, and pillow come with ComfyUI natively. The only new requirement is opencv-python (cv2).

---

👥 Credits & Acknowledgements

This pack was vibecoded with Qwen, Gemini and DeepSeek. Some of the nodes were inspired by Impact-Pack, rgthree-comfy, Chibi-Nodes, Fill-Nodes, art-venture, Geeky_AudioMixer, WhiteRabbit and afloy_audio_tools.

---

CHANGELOG

V1.1.0

• Added 6 new nodes: CB Audio Info, CB Audio Processor, CB Roll and Trim Audio Samples, CB Roll and Trim Video Frames, CB Nearest Resolution, CB Color Correct.
• Added requirements.txt to enable opencv-python (cv2) for CB Color Correct node.