My AI Team Now Has an Interface. All 12 Agents. Free.
Tom, creator of the iCola methodology, releases My PKA version 3 with a free interface (My PKA Cockpit) that provides 12 pre-built AI agents for productivity, note-taking, health tracking, document management, and task planning—all running locally in a folder structure accessible via Claude and a web interface.
Summary
Tom presents the evolution of his personal productivity system, explaining that for decades he's used various tools (note-taking apps, task managers, habit trackers) and extracted the best features from each to create an integrated workflow. He introduces My PKA version 3, a free scaffold that brings together all these capabilities in one local folder system that works with Claude and includes a new web-based interface called My PKA Cockpit.
The system includes 12 AI agents out of the box, along with previously paywalled expansion packs (app developer and designer packs) now included for free. The setup uses Markdown files, Git versioning, and a SQLite database to organize and connect information. Key features include fleeting notes with whiteboarding capabilities, document scanning and organization, health data integration, invoice tracking with automated reminders, calendar integration, and a planner application.
Tom walks through the installation process, demonstrating how to initialize the system by launching Claude in a terminal within the folder. The system is designed to be self-organizing—users interact with their AI team through natural conversation to set up connections to external services (like Todoist, health apps, and document scanners). The interface visualizes local folder contents in a dashboard showing today's actions, calendar, open invoices, documents, projects, goals, and habits.
Importantly, Tom emphasizes that the system is designed to be customizable and integrated into existing workflows. He demonstrates his personal setup, which contains 48 agents highly specialized to his specific work streams, and shows how users can request AI to build custom agents for their needs. He provides guidance on migrating existing scaffolds to the new version while avoiding breaking changes, emphasizing the importance of backing up folders before updates.
The video concludes with a demonstration of the Cockpit interface populated with demo data, showing features like connected invoices, serendipity functions (showing what happened on the same day in previous years), person databases with connections, recipe libraries, and health tracking with map visualization. Tom encourages users to join his community to share their setup experiences and announces upcoming courses including a new Claude Code course.
Key Insights
- The system uses a local folder structure with Markdown files, Git versioning, and a SQLite database as its foundation, allowing users to open and view everything in standard tools like Obsidian while maintaining efficient connections and visualization through the web interface.
- Tom shifted from providing a stripped-down, generalistic scaffold to releasing his full personal setup because he considers it more useful for others to have access to the same version he uses daily rather than a simplified alternative.
- The system uses natural language conversation with AI to establish connections to external services—users don't follow step-by-step instructions but instead tell the AI team what they want connected, and the AI provides guidance on obtaining API keys and integrating services.
- Tom's personal production setup contains 48 agents, but he only released 12 agents for free because most of his additional agents are highly specialized to his specific workflows and work streams, making them not generally applicable.
- The system is designed to integrate into existing workflows rather than replace them—users can ask AI to extract specific agents (like security, designer, and developer agents) from the new scaffold and compare interfaces to understand what functionality to bring over without breaking changes.
Topics
Transcript
[0:00] For decades, I've been using productivity tools all over the place. Note-taking apps, task managers, habit trackers, journaling applications, and they all helped me to become more productive. It was all about orchestrating these tools and combining them in the right way to take the most out of it. And this is what my iCola is all about and the iCola methodology helping you with exactly this. Based on all this experience, I knew what I wanted out of these tools. It was usually just one or two features that I [0:31] really appreciated. Example, in Hepta base, I had the whiteboarding tools that was great. In Tana, I had a outlining experience, what was great. In Day One,…
Full transcript available for MurmurCast members
Sign Up to AccessMore from ICOR with Tom | AI Productivity
You are using the wrong Claude for work! (Here is proof)
A video demonstrating that Claude Code is significantly more powerful than Claude Cowork for professional knowledge workers, showing how Claude Code better understands folder context, enforces safety rules, orchestrates sub-agents with different models, and creates persistent file outputs rather than temporary artifacts.
Obsidian is the tool I teach in. Keep the one you have.
The creator addresses criticism about switching to Obsidian after previously promoting other tools, clarifying that his tool-agnostic ICOR methodology remains unchanged and that Obsidian is simply an interface layer for his folder-based system. He emphasizes that the underlying folder structure, AI automation, and productivity principles are identical regardless of which tool visualizes the content.
Claude sucks. And here is why.
This video explores why Claude and other AI models produce inconsistent results, examining how model selection (Haiku vs Sonnet vs Opus vs Claude 3.5 Fable), effort levels, prompt clarity, and system context dramatically affect output quality. The speaker demonstrates these differences through a Venn diagram creation task and advocates for using organized folder systems with documented SOPs and code-based solutions to ensure consistent AI performance.
Claude can now prompt its own SESSIONS. I built the boss.
Claude now has the ability to interact with multiple sessions, allowing a streamlined orchestration of tasks through an agent called Larry. This integration can simplify task management and enhance productivity by enabling a single main session to oversee various sub-sessions or tasks.
I switched off Claude's memory. My folder does it better.
Tom explains why he disabled Claude's native memory feature and instead built a local folder-based system (myPKM) to maintain persistent context with AI agents. He shares an audit revealing systemic issues like contradictory documentation, forgotten rules, and parallel session conflicts, then outlines how to systematically fix these problems using the ICOR productivity methodology.