Building & APIs
Building GreenOS: what went wrong with my evidence corpus
The seven-minute research run, source-checking mistakes and broken restore that shaped how I am building a traceable cannabis research corpus.
The full storyqmd: local search for notes and knowledge bases
qmd is a command-line search engine for Markdown documents, knowledge bases and meeting notes, with keyword, vector and reranked search modes.
ACCESS AND VALUE
Free open-source software under the MIT licence
How I use qmd
I use qmd to search the Obsidian notes, project records and operating instructions in my Second Brain before work starts.
What makes it different?
It combines keyword search, vector search and reranking while keeping the indexed material on the machine running it.
Competitors
- Obsidian: A notes workspace rather than a command-line retrieval layer
- Khoj: A self-hosted assistant and search layer for personal knowledge
Product source checked 2026-09-22.
Who it is for
Searching local Markdown notes, project records and personal knowledge bases.
Who can claim it?
The source and installation instructions are publicly available. It runs locally with Node.js and downloads models for semantic search and reranking.
What the free software includes, and what costs extra
Local indexes and downloaded models use disk space. Semantic search and reranking need more processing than keyword search alone.
Billing and renewal
There is no recurring software charge; local hardware and storage remain separate.
- Tax basis
- Not applicable to the free software or plan
- Card required
- Provider’s signup terms
- Expiry
- No fixed availability end date is listed.
- Public research date
- 22 Sept 2026
- Source checked
- 22 Sept 2026
Further official evidence
Source and corrections
Guides featuring qmd
Work & productivity
How I made my Second Brain useful with qmd
Local search for my Obsidian notes, the stale instructions it uncovered, and the work that made previous decisions useful to me and my coding agents.
The full story
AI & digital tools
How I built two publishing pipelines without making both sites the same
Swarm Signal and MCPH share a process, but serve very different readers. The broken audio handoff, stale pages and writing fixes that taught me where to draw the line.
The full story