← Back Home

The MJ mentality, Part II finding your edge

Daily Tasks

Meditations

0800 AM

Woke up at 3:30 am today, meditated. Still mentally exhausted asf. Went to bed, slept until 7:30 AM and got in later than expected.

Reoccuring thoughts - palantirization of software services x enable optimal individual freedom/balance. AI shouldn’t be leading to more work and burn out, it should enable LESS human draining work, and give back time and energy.

The product engineering is a sore spot right now, I don’t quite want to work. The dreadful dilemma of execution or rest/replenishment. I have taken yesterday off to mentally recalibrate, this week has been more demanding mentally.

Another solution is minimizing resistence in workflows - ie working smarter with 1) keyboard shortcuts - reduces 20% cognitive energy 2) optimized filing 50% less mental energy 3) scheduling physical tasks during breaks to efficiently harness natural energy states.

Mental energy arises from context switching unexpectedly.

The default mode is to be doing nothing, recovering and recharging - not overworking. Timeblock strictly engineering/energy intensive tasks and be ruthlessly deliberate about their strategic value and alignment.

Tighen iteration loops. Timeblock to constrain time, pack more reps into a set time period. More reps = more knowledge -> more data for informed realtime decision making.

Workspace

— 1640 PM — Filesystem MCP Server: Allows LLMs to interact with your local file system, including reading, writing, and managing files. This is fundamental for agents needing access to local documents or code.

GitHub MCP Server: Facilitates interaction with GitHub repositories, letting AI agents perform tasks like checking issues, reviewing pull requests, and analyzing commits. This is especially useful for developers.

Playwright MCP Server: Enables browser automation, empowering AI agents to interact with web pages, perform scraping, and automate web-based workflows. This server is popular with 12K stars on GitHub

— 1620 PM — Building by hand is too slow, I need an MCP co-pilot to build.

My painpoint is that it takes alot of mental energy to learn the exact format/templates for each tool, and integrating the APIs and keeping variable names/data structures consistent is difficult.

Is there an almost “meta-agent” which can be my engineer and figure these integrations out? Are there any MCPs which can build me these workflows incorporating different tools/stacks?

— 1615 PM —

How do I use page designer block to generate a custom pdf output, with data values inserted?

Figure out how to architect automation workflows. What are the user inputs? What are the outputs?

I want the following agent: Input: copy/paste txt portion, partially in json, give request to execute (addVehicle, leaveMessage) Output: create PDF with forms filled.

Airtable Community - Printing Copy of Completed Form

Use Fill Out as 3rd party

— 1100 AM — Can you implement a .json input feature, where I can copy/paste json and it generates an entry?

For example, this input I’m pasting in:

9485393 : { n : Guillermo Ganos e : Guillermo@YAHOO.COM r : addVehicle() { vin: SMFVC4SD4SU053726 purDate: 09/16/2025 }

c : 

}

where n : name, e : email, r : request, vin : VIN, purDate : purchase date. Also, The inital number is the primary key, which is policyId

NCFB Acceptable Use:

This policy does not cover traditional AI tools (e.g., Google’s search algorithm), which refers to systems designed to respond to a particular set of inputs and perform a specific task. Traditional AI follows predefined rules and does not possess the ability to generate new content or adapt to new situations without explicit human intervention.

Data Hygeine practices

I also deidentify data - using dummy data names, no acutal sensitive personal idenifers like policy number, DOB, etc…

I plan on encrypting these before entering them into the algorithm, so its not ingesting sensitive proprietary data.

I also have separate professional workspaces and personal. I practice good data hygeine.

What other data compliant safeguards or practices should I consider?