Files live where you decideβoffline, encrypted drives, or private repos. No opaque databases or vendor lock-in. Just files.
Small, purpose-built docs that let you mix and match context packets for any AI conversation or Agent task.
Optimized for machine reasoning and human browsing. If you can't read it, you can't trust what the AI knows.
MCL bridges the gap between personal note-taking and AI infrastructure.
(Obsidian, Notion, PARA)
(Vector DBs, MemGPT)
A standardized yet flexible schema for your digital life.
MyContextLibrary/ βββ agents/ # π€ Agent configs, prompts & memory βββ export/ # π€ Generated PDFs & reports (Outputs) βββ import/ # π₯ Raw ingest folder (Inputs) βββ library/ # π§ The Core Knowledge Base β βββ me/ # - Brand, Bio, Facts β βββ projects/ # - Active endeavors β βββ work/ # - Companies & Products βββ playbooks/ # π Reusable templates & guides βββ routines/ # π Recurring workflows βββ tools/ # π οΈ Automation scripts βββ settings/ # βοΈ System configs
You can start building your library in less than 5 minutes.
Simply create a folder named MyContextLibrary on your disk and add the core subfolders: library, me, and settings.
Create library/me/fast_facts.md. Add the Minimum Viable Header (MVH):
Don't over-engineer. Add projects/ when you start a new project. Add playbooks/ when you find yourself repeating a task. Let the structure evolve with you.
Integrate your library with your favorite AI tool or agent like Claude Code, Gemini CLI, or others. Don't be afraid to experiment!