MyPKA
A local-first markdown folder plus a six-role AI team, run on the ICOR methodology and readable by any LLM you choose.
MyPKA is the strongest choice here if data ownership and model independence matter more to you than a finished interface. The six role contracts plus SOP-001 and SOP-002 give an LLM something Notion AI and Obsidian-plus-plugins do not: a defined job per specialist and a procedure for adding one. The trade-off is that it is a folder scaffold, not an app — you supply the LLM client and the markdown comfort, and you handle backup and sharing yourself. Pick it if you are method-driven and want your context to survive whichever model ships next; skip it if you need a GUI, real-time team collaboration, or a cloud-first sync layer.
Verified 7d ago · liveness 57/100 · cite: rightaichoice.com/tools/mypka
- Knowledge workers running a formal method such as GTD, PARA or ICOR
- AI power users who want model-agnostic persistent memory
- Solopreneurs needing structured AI collaboration without a platform account
- People who prioritise data ownership and local-first storage
- Users who want a polished GUI rather than file and folder interaction
- Teams needing built-in real-time collaboration
- Anyone seeking a cloud-first sync solution out of the box
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip MyPKA if you want a finished app with cloud sync and a GUI, rather than a local markdown folder you point Claude, Gemini or Codex at yourself.
The open-source folder core is separate from the myicor.com teaching side, so the ICOR Journey courses and weekly live sessions with Tom and Paco are a second spend on top of running the tool.
There is no published plan table in this scrape, so treat MyPKA as a free-to-run open-source folder plus an optional paid teaching membership at myicor.com. That combination lands cheaper than a per-seat knowledge platform once your team grows, but it costs you setup time and a separate LLM provider bill that seat-based tools fold into one invoice.
In short
MyPKA — A local-first markdown folder plus a six-role AI team, run on the ICOR methodology and readable by any LLM you choose. Best for Knowledge workers running a formal method such as GTD, PARA or ICOR, AI power users who want model-agnostic persistent memory, Solopreneurs needing structured AI collaboration without a platform account. Free to start; paid plans from $79/mo.
What people actually say about MyPKA — is it worth it?
We ran a structured research pass across product reviews, community discussions, and post-purchase forum threads to surface the patterns vendors won't publish themselves. Below: the recurring strengths, the hidden costs people mention most, and the cohort that consistently regrets adopting this tool.
6 mentions across 1 source (GitHub) · researched Jul 3, 2026.
Average across the 1 source that answered — each source counts once, not each post.
- +Local-first design gives full data ownership—no cloud or login needed.
- +Plain markdown files are fully portable across any LLM or editor.
- +Open source under CC BY-NC-SA 4.0 encourages transparency and forking.
- +Six AI specialists with defined roles structure complex knowledge tasks.
- +Persistent context across sessions via journals and contracts.
- −Multiple critical bugs (white-screen, MCP crash) break core functionality.
- −Upgrade mechanism points to a broken URL—can't update easily.
- −Light mode UI has rendering issues that hinder readability.
- −Planning page layout is disoriented on small screens.
- −Community feedback is sparse—only six GitHub posts available.
- • Self-hosting may incur cloud storage or compute costs if users choose to sync via third-party services
Viability Score
How well maintained and how widely used is MyPKA? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: October 2026
How we score →Key Features
- Local-first folder of plain markdown files you own outright
- Six named AI specialists: Orchestrator, Talent Acquisition, Research, Capture, Connection, Database Architect
- Per-role contracts and journals that hold context between sessions
- Model-agnostic — runs against Claude, Gemini or Codex
- MCP server so an AI client can read the knowledge base directly
- Hiring procedure SOP-001 to define and add a new specialist role
- SQLite migration support via SOP-002
- Research protocol that reports findings with confidence levels
- PKM routing map that categorises each input as it lands
- Weekly review automation for the Refine step of the loop
- Deliverable generation in structured output
- No app, login or cloud account required to run it
- Built on the four-move ICOR loop: Input, Control, Output, Refine
About MyPKA
MyPKA is the plain-folder half of the ICOR methodology taught at myicor.com: instead of an app, you get a folder of markdown files on your own disk, holding your notes, projects and decisions in a structure any AI can read. Inside it sits a scaffolded AI team of six named specialists — Orchestrator, Talent Acquisition, Research, Capture, Connection and Database Architect — each with its own role contract and journal, plus procedures such as SOP-001 for hiring a new specialist and SOP-002 for migrating content into SQLite. Because the files are plain markdown and the team is model-agnostic, the same context persists whether you point Claude, Gemini or Codex at the folder, and nothing is locked to a vendor account. It also exposes an MCP server so an AI client can read the knowledge base directly. This is aimed at knowledge workers who already run a method (GTD, PARA, ICOR) and want their AI to work on top of ordered input rather than a chat window, not at people who want a polished GUI or built-in cloud sync.
Behind the Verdict
The interesting thing about MyPKA is what it refuses to be. It is not a note app with an AI button bolted on — it is a folder of plain markdown files with a named team of AI specialists standing around it, and that inversion changes what you get for your money. On the myicor.com site the ICOR loop is stated plainly: Input, Control, Output, Refine, with the explicit position that AI comes last, once notes, tasks and projects are already one ordered system. MyPKA is the concrete implementation of that "one folder" path — the site calls it out as the radical-simplicity option where the whole methodology lives in local files any AI can read, naming Claude, Gemini and Codex as the readers. Strengths. First, the team has contracts, not vibes: the Orchestrator, Talent Acquisition, Research, Capture, Connection and Database Architect roles each carry a defined remit and a journal, so context carries between sessions instead of being re-explained. Second, SOP-001 makes the team extensible — you run the hiring procedure to define a new role and generate its contract rather than hand-rolling prompts. Third, SOP-002 provides a SQLite migration path, which is the right escape hatch for when a folder stops being enough. Fourth, the MCP server means an AI client reads the knowledge base as a system, which one community member on the site described as giving the AI "context on what I have written." Fifth, everything is portable: no app, no login, no cloud account standing between you and your own material. Weaknesses. This is genuinely more demanding than a normal SaaS signup. The filesystem is the interface, and if markdown and folder hierarchies do not sit naturally with you, the setup cost will outweigh the payoff. There is no built-in cloud sync — the site is candid that the local folder model puts backup in your hands — and no real-time collaboration for teams. The membership side of myicor.com (the Journey courses, the weekly live sessions with Tom and Paco, the community feed) is how you learn the method, and it is billed separately from the open-source core, so budget for both the tool and the teaching if you are starting cold. Where it fits. Solo knowledge workers, consultants and small operators who already run a methodology and want their AI working on ordered ground. Where it does not: anyone who needs a GUI-first experience, a cloud-first sync solution, or a team workspace with concurrent editing — for those, Notion AI or Heptabase will feel far less like homework on day one.
Researching MyPKA? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Real-world workflow fit
Concrete scenarios for the personas MyPKA actually fits — and what changes day-one when you adopt it.
Capture client notes as plain markdown in the local folder during the week, then have the Orchestrator and Research specialists pull them into a Monday plan, with the Research role tagging each finding with a confidence level.
Outcome: One ordered folder that any LLM client can read on Monday morning, with no re-prompting of past context and no vendor account holding your client material.
Point each model at the same MCP-connected folder so the Capture and Connection specialists file new material into the same structure regardless of which model is answering that day.
Outcome: Model choice becomes interchangeable — your knowledge base keeps compounding in one place instead of fragmenting across three chat histories.
Run SOP-001 with the Talent Acquisition specialist to define a missing role and generate its contract, then let it file into the SQLite-backed store set up through SOP-002.
Outcome: The team grows on a defined procedure rather than ad-hoc prompt tinkering, and the larger store keeps queries fast as the base grows.
Use Cases
- Persist research context across Claude and Gemini sessions without re-prompting from scratch.
- Run a named Research specialist with confidence levels so you can see how firm each finding is.
- Onboard a new AI role using SOP-001 instead of writing the contract prompt by hand.
- Move a growing knowledge base into SQLite when the folder alone stops scaling.
- Keep project goals and decisions in a version-controlled folder any AI client can read.
- Have the Orchestrator pull notes from each specialist into a single weekly review.
Models Under the Hood
as of 2026-10-11
Limitations
- MyPKA is a methodology and folder scaffold, not a standalone application — it needs an LLM client and genuine comfort with markdown and a filesystem.
- It maps the tools you already own rather than replacing them.
- There is no built-in cloud sync, so backup and sharing are manual jobs you own.
- Non-technical users will find first setup demanding.
- The teaching side of myicor.com (the ICOR Journey courses, weekly live sessions and community) is separate from the open-source core, so the method and the tool are bought and learned on different tracks.
as of 2026-10-04
Verification history
We have re-verified MyPKA 8 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — re-checked, vendor evidence unchanged
- — re-checked, vendor evidence unchanged
- — re-checked, vendor evidence unchanged
- — re-checked, vendor evidence unchanged
- — re-checked, vendor evidence unchanged
- — re-checked, vendor evidence unchanged
Showing the 6 most recent of 8 verification passes.
Free to cite with attribution — this page re-verifies continuously.
12-month cost
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
Where the pricing makes sense
The company stage and team size where MyPKA's pricing actually pencils out — and where peers do it cheaper.
There is no published plan table in this scrape, so treat MyPKA as a free-to-run open-source folder plus an optional paid teaching membership at myicor.com. That combination lands cheaper than a per-seat knowledge platform once your team grows, but it costs you setup time and a separate LLM provider bill that seat-based tools fold into one invoice.
Setup time & first value
How long it actually takes to get something useful out of MyPKA — broken out by persona, not the marketing-page minute.
For a method-driven consultant already comfortable in markdown: roughly an afternoon to clone the folder, wire an MCP-capable AI client and run the first loop. For a power user mixing Claude, Gemini and Codex: a couple of hours to point each client at the same folder. If you are new to markdown and folder hierarchies, budget several evenings, since the ICOR Journey courses that teach the loop are
Switching to or from MyPKA
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From Notion or Obsidian: export your notes to markdown, drop them into the MyPKA folder and let the Capture and Connection specialists re-file them against the PKM routing map.
- →From a chat-only AI workflow: move your recurring context out of chat history and into the folder, so each specialist reads from files instead of a pasted transcript.
- →From a folder that has outgrown itself: run SOP-002 to migrate the knowledge base into SQLite once plain markdown lookups slow down.
- ↗To Obsidian: point a vault at the same markdown folder, since the files are already plain and portable.
- ↗To a SQLite-backed store: export through SOP-002 if you want structured queries instead of flat files.
- ↗To any other AI tool: hand over the folder directly — it is plain markdown with no proprietary format to unpick.
Resources & Guides
Tutorials & Learning
YouTube returned 6 videos for “MyPKA”, and we withheld 6: 6 could not be judged, because “MyPKA” is a single word that other videos use for other things. We are showing none, because we could not prove any of them are about MyPKA.
Official links
Tools that pair well with MyPKA
Common stack mates teams adopt alongside MyPKA, with the specific reason each pairing earns its keep.
remio
Local-first AI agent that turns your files, meetings, and research into finished decks, spreadsheets, and reports.
Reflect AI
Open-source, markdown-first note-taking app with daily notes, backlinks, and optional AI on your own API key.
Sakana AI
Sakana AI builds Japanese-specialised LLMs and multi-model orchestration for finance, defense and intelligence teams that must keep data in
Featured Head-to-Head Comparisons
Mypka vs Praktika
Praktika is a mobile app for language learners who want to practice speaking with AI tutors. MyPKA is a desktop system for knowledge workers who need a persistent, local-first AI team. They serve entirely different needs—choose Praktika for language fluency practice, MyPKA for organizing and retrieving knowledge across sessions.
Mypka vs Surge Ai
Choose MyPKA if you want a free, open-source, local-first knowledge management system with persistent AI context and zero vendor lock-in. Choose Surge AI if you need expert human feedback for RLHF, red teaming, or complex benchmarks—especially if you're a frontier AI lab or enterprise building advanced models. They serve completely different needs; MyPKA is for personal productivity, Surge AI is for AI training and evaluation.
Alternatives to MyPKA
View all
remio
Local-first AI agent that turns your files, meetings, and research into finished decks, spreadsheets, and reports.
Reflect AI
Open-source, markdown-first note-taking app with daily notes, backlinks, and optional AI on your own API key.
Frequently Asked Questions
Best-of guides
Used MyPKA? Help shape our editorial sentiment research.