Local tool condenses PDFs and docs into short summaries and goal-focused reading

The open-source project tldr-everything runs entirely on a local machine and turns PDFs, Markdown files, or pasted text into three summary layers: a 30-second overview, a 3-minute explanation, and page-cited detailed notes. A second mode lets users state a goal, such as deploying with Docker on a VPS, and returns only the sections relevant to it, scored for relevance. It uses the Qwen2.5-3B-Instruct model through Ollama, which needs roughly 8 GB of RAM and can run CPU-only. The pipeline splits work into single-task calls and verifies outputs in Python, dropping fenced commands absent from the source and flagging unverified inline ones. The code is MIT-licensed, though the default model weights carry separate licensing terms.