Use MarkDone as a local LLM context builder and context.md generator for documents, notes, configs, and exports. Drop PDF, DOCX, Markdown, JSON, YAML, CSV, or text files, then build one structured Markdown context pack for any LLM without uploading source files.
context.md generatorTurn mixed files into a structured Markdown handoff with clear source boundaries and metadata.
Secret-awareDetect and optionally redact common keys, tokens, private keys, and password-like values before pasting.
Private by designPDF extraction, DOCX parsing, scanning, and token counting run in the browser.
How to build an LLM context pack
- Drop PDFs, DOCX files, Markdown, configs, CSVs, JSON, YAML, or text.
- Keep Redact secrets enabled unless you need exact raw text.
- Copy the Markdown for any LLM or download context.md for reuse.
When an AI context builder helps
Use it when a task needs more than one file: product notes, support logs, repository docs, API payloads, research PDFs, customer feedback, or deployment configs. The output keeps the source material inspectable before it enters an LLM workflow.
Why a generic LLM export
The output is plain Markdown with source markers, metadata, and warnings. It works across chat apps, coding agents, local models, and API workflows, so provider-specific copy buttons would add noise without adding value.
What stays local
Your documents, extracted text, token counts, and secret warnings stay on your device. MarkDone does not send file contents to a backend, model provider, or third-party conversion service.
PDF and DOCX to LLM context
Readable PDFs and Word documents can be converted into source sections alongside plain-text files. That makes it easier to combine specs, reports, notes, and configuration files into one prompt-ready context pack.
Context pack vs token counter
A token counter only estimates size. A context pack preserves file names, source boundaries, metadata, warning counts, and redacted content so the final Markdown can be reviewed before it is copied into an LLM.