The AI-CODE Workflow
We have upgraded the PKM from a static repository into a Personal AI Infrastructure (PAI). This shift moves us from manual filing to dynamic data pipelines, using Tiago Forte’s AI-centric CODE and PARA frameworks.
1. The Architecture: PARA as Context Bundles
In this system, PARA folders are no longer just storage; they are Context Bundles. When working on a task, PAI loads only the relevant folder to prevent “context poisoning.”
graph TD A[Capture: Alpha Data] -->|Inbox| B(Distill: Master Prompt) B -->|Organize| C{PARA Context Bundles} C -->|Project| D[The Strategist Persona] C -->|Area| E[The Architect Persona] C -->|Resources| F[The Librarian Persona] D & E & F -->|Express| G[AI-Enhanced Output]
2. The Board of Advisors
We’ve established specialized AI personas (Agents) tied to specific PARA categories:
| Persona | PARA Context | Specialty |
|---|---|---|
| The Strategist | the-strategic-bridge/ | Strategic frameworks, long-form writing, and decision science. |
| The Architect | pattern-languages/ | Systems thinking, modular design, and structural PKM integrity. |
| The Librarian | dispatches/ (Inbox) | Distillation, data cleaning, and Master Prompt generation. |
3. The Distillation Engine
The “Distill” phase is now automated. Using the distill_master_prompt.md template, PAI takes raw “Alpha” data—transcripts, niche captures, and offline conversations—and transforms them into operational manuals.
The Protocol:
- Capture: Drop raw
.mdfiles intocontent/dispatches/. - Distill: Ask PAI to “Distill this dispatch into a Master Prompt.”
- Organize: Move the distilled manual into the relevant PARA Project or Area.
- Express: Invoke the corresponding Agent from the Board of Advisors to iterate on the output.
4. Why This Works
- Clean Data: We focus on “Alpha” data—info the AI doesn’t already have in its training set.
- Precision: Context Bundles keep the AI’s focus sharp.
- Velocity: Manual organization is replaced by agentic orchestration.
Generated by PAI in collaboration with Garth.