Archive

personal note: i am not using github for login, but gmail account instead.

Leased this car in November 2024 from Tomwood Vinfast. I had originally called them to enquire about Subaru's EV, and the salesman convinced me to take a look at this car insted for a better deal. And I did.

• Zero down.
• $343/Mo
• Insurance $855/6months = $142.5 = $485.5 (In comparison, my 2025 Mazda CX90 costs $482/6months )

If I drive thousand miles per month, I'm spending almost $2/mile, accounting for most of my EV charging happens at work.

#courseware

iximiuz Labs - Indie Learning Platform to Master Server Side Craft, via X/DeepKnowledgeU. See How (and Why) to Use containerd from the Command Line (course) #kubernetes

About doing citations correctly...

• See Citation File Format (.cff) on github. Github, Zotero all support this yaml format.

Sources on keeping up with AI.

This is a good start as of July 2025 - How I keep up with AI progress (and why you must too) from Nilenso blog, since I use many of the same.

LLM Inference in Production by BentoML; "core concepts and performance metrics, to optimization techniques and operation best practices"

Agent-to-Agent - A2A enables multiple agents to communicate, coordinate, and delegate tasks to each other, even if they come from different vendors or frameworks.

Key Components of A2A:

• Agent Card -- JSON File. lists agent's capabilities, endpoint URL, and authentication method.
• Task -- Unit of Work. Task has lifecycle: submitted, working, completed of failed
• Message -- Structured communication between agents, exchanged within the context of a Task
• Part -- blocks within messages
• Artifact -- final result of a Completed Task, returned to client

Read more here: MCP vs A2A Clearly Explained

See also MCP

Model Context Protocol -  is designed to let a single agent securely access tools, APIs, and data sources in a standardized way.

On average, RAG, MCP, and NLWeb reach about 0.75 F1, while HTML stays near 0.67 and misses more. HTML agents also burn more compute, needing about 3x more tokens and 5x more time per task than others. So the study recommends RAG agents, often with GPT5 or cheaper GPT5 mini, and keeping HTML browsing only as backup. -- (Steiner et al. 2025)

--

Let your Coding Agent debug your browser session with Chrome DevTools MCP  |  Blog  |  Chrome for Developers

• "Let your AI agent see and interact with your live Chrome session" -- GitHub - pasky/chrome-cdp-skill: Give your AI agent access to your live Chrome session — works out of the box, connects to tabs you already have open · GitHub

See also A2A

References

• Steiner, Peeters, Bizer (2025). MCP vs RAG vs Nlweb vs HTML: A Comparison of the Effectiveness and Efficiency of Different Agent Interfaces to the Web (Technical Report).

Gemini CLI is the equivalent of Claude Code.

npm install -g @google/gemini-cli

My LinkedIn page is: linkedin.com/in/btbytes/