@openclawzhangchong

moondream-vision-zc wraps a local ollama-deployed moondream model for image understanding, accepting file paths or embedded images in openclaw chat and returning natural language descriptions. integrates with gpt-oss-120b for downstream reasoning tasks.

douyindownloadwhisper extracts watermark-free douyin videos and transcribes speech to text using local whisper, then auto-segments transcripts via llm. windows-only, command-line driven.

qq-mail-read-send enables imap-based qq mailbox reading and smtp sending with multi-encoding support. lacks explicit error handling, decision logic, and output contracts despite providing functional code snippets.

doubao-opencli wraps the doubao ai assistant via opencli browser bridge, automating chat, podcast generation (text/url/pdf input with ad trimming), ppt generation, image synthesis, meeting summaries, and batch operations. requires edge login and opencli installation.

html-ppt-zc generates pptx and optional html presentations from title and body text, splitting content into slides via python-pptx. supports basic styling and browser-viewable output.

zc-douyin-xiazai-txt downloads watermark-free douyin videos, extracts audio via ffmpeg, transcribes with whisper, and optionally segments transcripts using an llm for improved readability.

pdf-ocr-zc wraps ocrmypdf and tesseract to batch-process scanned pdfs, adding searchable text layers and optionally exporting markdown or plaintext. targets teacher agents automating textbook ingestion.

Performs basic FTP operations like list, upload, download, delete, and rename using stored or provided credentials with Node.js basic-ftp.

Extract text from images via OpenRouter API using Baidu Qianfan OCR model, supporting URLs and local files with customizable prompts.

自动将PPT或文本脚本转换为带讲解配音的演示视频，支持多种幻灯片模板和中英文配音。