Loading listing…
Synthadoc
Security scan passedPull text and content out of documents, slides, sheets, PDFs, images, web pages, YouTube and AI sessions for your agent.
Pack · 9 skillsby axoviq-ai· ProductivityUnclaimed · Yours? Claim it
Preview
4 images
What Synthadoc does
About
Synthadoc is axoviq-ai's open-source LLM wiki engine: it reads your raw source documents and synthesizes them into a structured, cross-referenced Markdown wiki. This pack bundles its ingestion Agent Skills, which pull text and content out of many file types so your agent can work with them. The skills in the pack - docx: extract paragraph text from Word documents - pdf: extract text from PDFs, with a fallback parser for CJK fonts - pptx: extract slide text and speaker notes from PowerPoint - xlsx: extract data from Excel and CSV files - image: extract text from images using a vision model - markdown: ingest Markdown and plain-text files - url: fetch and extract text from a web page - youtube: pull transcripts from YouTube videos - session: extract conversation turns from AI session history (.jsonl) How it installs - Install the skills with npx skills add axoviq-ai/synthadoc. Each skill is a small Python script and names the library it needs (for example python-docx, pypdf or python-pptx). - The full Synthadoc wiki engine, which builds the cross-referenced wiki, detects contradictions, flags orphan pages and cites sources, runs separately (Node, with a Docker option) and stores its output as local Markdown for Obsidian or any wiki-compliant tool. Good to know - The extraction skills read content; they don't change your files. - Open source under the AGPL-3.0 license.
More from Productivity
Skills people install alongside Synthadoc.
