BD | Blog

Figma use of Agents

Fri, 21 Aug 2026 00:00:00 GMT

Thoughtful blogs from figma engineering:

AI Local Apple Silicon project

Tue, 04 Aug 2026 00:00:00 GMT

Always on the lookout for some AI local options and this project got my attention:

AI SDLC

Thu, 30 Jul 2026 00:00:00 GMT

Another great Beyond Coding interview AWS Veteran: The New Software Development Life Cycle - YouTube.

The most practical takeaway is Lessa’s model cost discipline. After a single refactor cost 200 million tokens because Opus was “overthinking” every line, he split the workflow into three tiers:

Tier 1 — SOTA models (Opus, Sonnet): planning, spec writing, architecture decisions only Tier 2 — Mid-tier models: 70–80% of implementation work Tier 3 — Cheap models: code review, linting, adversarial checks

Product loop

discovery Human
whiteboard Human
/roadmap Human+AI
/roadmap —sync
/new-work Human+AI
- dev loop

Dev loop (Heitor Lessa)

/opsx:explore - human+agent sota
/opsx:plan - human+agent sota
/opsx:apply - agent
refinement - human+agent
merge-check - agent+gate
review & merge - human
/retro - human. workflow loop

Taming mono-repo

Fri, 24 Jul 2026 00:00:00 GMT

2026-07-24

Setting up a new project(s) and I’m still looking for improvements that give me flexibility.

In particular, wanting to use deno more and do use expo for mobile clients.

Inspired by this article, feel this will be very helpful:

https://metabureau.com.au/blog/deno-pnpm-moonrepo-monorepo

Grok impresses on a simple task.

Sun, 19 Jul 2026 00:00:00 GMT

See the mvp here: Reset Tracker

Kanata FTW!

Sun, 05 Jul 2026 00:00:00 GMT

Problem: Alt. keyboard layout issue differences between my exp. and laptop internal keeb.

I’m building my new custom low-profile 38-key (3x6) -2 +3 thumbs - but I need to solve a problem I’ll hit.

My Piantor is configured with Vial / QMK with

Colemak-DH
Home row mods
symbol layers
combo keys

The mac internal keyboard has no layers and minimal special functions.

Colemak-DH ANSI installed as input source
CapsLock as LeftCtrl per ModifierKeys
HyperKey allows Super ROpt / RMet
Aerospace keyboard shortcuts dont work when i switch, it needs stable input source.

The workaround so far: switch input source to Colemak-DH when on laptop with internal keyboard and to Australia when using the Piantor split keeb. Lots of down-sides:

no layers or combos on mac internal keyboard
Aerospace shortcuts I had to use the qwerty location always.

Action / Research

I’ve researched most of this before, I’ve tried Karabiner-Elements but I didn’t want the complexity of the kernal extension. I’d come across kanata, saw it needed that as a dependency and wrote it off.

A chance Gemini / Google AI mode question put me onto the path that let me rediscover kanata and how it can depend on the karabiner device-kit architecture.

Result

The result is I have a working base kanata configuration with qwerty (for others/troubleshooting) and colemak-dh. Keeps my Aerospace shortcuts same for either layout. I’ll continue to add the additional layers and combo keys I have from vial.

This enables me to try Sturdy / Magic Sturdy regularly till proficient - when working

It’s not 100% smooth, I’ll keep testing out and find any other issues. E.g My Arc shortcuts aren’t working.

Learnings

Continual probing of AI helps us explore the problem space and solution space quickly to eliminate sub-par paths and save time. I re-discovered a major problem worth solving (the issue was back of mind) it can help find the path to follow before commiting much time.

Dwarf Star 4

Sat, 04 Jul 2026 00:00:00 GMT

GitHub - antirez/ds4: DeepSeek 4 Flash and PRO local inference engine for Metal, CUDA and ROCm · GitHub

R&D / AI: mini-rag evaluation of cross-encoder efficiency

Mon, 29 Jun 2026 00:00:00 GMT

R&D / AI: Mini-RAG Cross-Encoder Efficiency Evaluation

Plenty of strong RAG content exists on papers and evaluation (e.g., Miraflow Guide).

Forked florianbuetow/mini-rag for local deep dive: how much efficiency gain from the cross-encoder?

Added token test in bdargan/mini-rag + bdargan/mini-rag-runner container. (Output posted below showing savings.)

See rag_token_test.py.

[TOKEN BENCHMARK REPORT]
  • Baseline Prompt Tokens (K=10): 308
  • Reranked Prompt Tokens (K=3):  112
  • Absolute Token Savings:         196
  • Efficiency Gains:               63.64%
PASSED

Key takeaways

Code from Gemini impressed again.
Follow-up: evaluate with Ragas

AWS lambda microvm

Wed, 24 Jun 2026 00:00:00 GMT

Well this will be useful. AWS introduces Lambda MicroVMs for isolated execution of user and AI-generated code - AWS

Priorities

Fri, 19 Jun 2026 00:00:00 GMT

Do you also struggle with those “frog” tasks that are important, but somehow you procrastinate enough to always put them off, till I probably used more time avoiding than it actually takes to complete. :facepalm:

Well I was trying a few ideas, beyond Eisenhower matrix, and came across the low energy task model and the Friction-Sized framework.

Google AI suggested a method to combine, so if identify tasks in md with X.Y.Z, it offered variations depending on tools used and search/filter needs see query

X=Importance (1-2) Y=Urgency (1-2) Z=Friction (1-4)

Z is 1-4 quadrants even though my rule of thumb is to batch the little things for a low energy time. I may sometimes hit those < 2mins immediately and skip the overhead of writing it down, but how many things reallly are 2mins and don’t spin you off distracted elsewhere.

An interesting experiment to help me match my mental energy with different tasks.

The “Friction-Sized” Framework

Because these tasks have deadlines and evoke reluctance, traditional advice like “Eat the Frog” actually backfires. When your mental energy is depleted, staring at a giant “frog” causes Executive Dysfunction—making you freeze and do nothing.To solve this, we must mathematically re-engineer how we measure task weight. Instead of measuring by Time (Duration), we must measure by Friction (Resistance to Start).

Total Effort = Friction (Energy to Start) + Drag (Energy to Complete)

1. The Friction Matrix

Use this 4-quadrant system to categorize both your work and personal life tasks. The goal is to isolate the “Sticky Notes”—tasks that take very little time to complete but have massive mental resistance.

	Low Drag (Takes <15 Mins)	High Drag (Takes >15 Mins)
High Friction(Reluctant to Start)	🛑 The Sticky Notes • Paying a bill / Renewing a license • Sending a “difficult” but short email • Booking a dentist appointment	🐸 The True Frogs • Tax returns • Writing a project proposal • Planning a major trip
Low Friction(Easy to Start)	⚡ Quick Wins • Deleting spam / Sorting mail • Putting away dishes • Approving a simple request	🏃‍♂️ The Momentum Builders • Organizing a closet / Desk • Exercising / Walking • Data entry while listening to music2.

2. The 3 Rules of Engagement

To defeat the “Sticky Notes” and “True Frogs” before their deadlines hit, apply these rules:

Rule 1: The “Micro-Leap” (For High Friction, Low Drag)

Never write the actual task on your to-do list. Write the absolute first physical action.

Instead of: “Pay car registration” (High Friction)
Write: “Open laptop and pull up registration email” (Zero Friction)
The Logic: Once the laptop is open and the email is in front of you, the friction drops to zero, and the 5-minute task gets done.

Rule 2: The 10-Minute “Micro-Block” (For True Frogs)

Do not try to “Eat the Frog” all at once. Set a timer for exactly 10 minutes. Tell yourself: “I am allowed to completely stop and quit after 10 minutes.”

The Logic: The hardest part of finite mental energy is the anticipation of the effort. A 10-minute boundary tricks your brain into starting. Usually, momentum takes over.

Rule 3: The Work-Life Hybrid Sweep

Never separate your personal and work to-do lists when energy is low. Your brain draws from the exact same central tank of mental energy for both. Batch your “Sticky Notes” (both work and personal) into a single 30-minute window right before lunch or at the very end of the day.

AI guardrails

Wed, 17 Jun 2026 00:00:00 GMT

Interesting interview on AI guardrails through the use of tests with semgrep for various languages. Why The Best Software Engineers Are Solving Code Review Bottlenecks Now - YouTube

GitHub - florianbuetow/ai-guardrails: Templates for working with AI · GitHub

He has other interesting projects with mini-rag and x-rag

Excalisave for CodeBerg - try out grok build 0.1

Mon, 15 Jun 2026 00:00:00 GMT

I quite like this chrome extension - https://github.com/atharvakadlag/excalisave plugin.

Only problem is the lack of alternative providers.

While having some time to kill (watching a youth volleyball grand final!). I had plenty of time between points to prompt.

I’m trying out different models and providers, so I gave Grok 4.1 with opencode.ai.

Couple more hours Monday morning to get it working. I was hoping to just be in product mode, but I had to go to the codeberg api docs to identify a cause of failure on document create was the incorrect HTTP method. Kinda annoying, it is clearly specified in the swagger docs.

Last iteration it needed network robustness, exponential backoff etc. and a more visual feedback.

It’s cost about $15 in tokens, which idk, I kinda expected it to have been < $10 AUD, going to be more < $20 ($USD 13). Either way it is pretty cheap - I’ll be using this weekly if not daily.

Excalisave fork https://github.com/bdargan/excalisave.

Needed Attention

Fri, 12 Jun 2026 00:00:00 GMT

Google’s response to prompt “climbing needed attention concept”:

In climbing, “attention” is the mental management of your focus, energy, and perception. Moving from merely “getting up a wall” to expert climbing requires intentionally shifting your attention between the environment, the mechanics of your body, and your subconscious movement. [1, 2, 3, 4, 5]

The Core Concept: Where You Place Your Attention

Effective climbing breaks down into a spectrum of focus: [3, 6]

Outcome Focus: Concentrating on the end goal (e.g., “I need to reach the top”). While necessary, this drains mental energy and causes stress if overused mid-climb. [3, 7, 8, 9]
Process Focus / Perception: Paying attention to what your body is doing in the present moment, such as the tension in your core, your breathing, and the precise placement of your hands and feet. [3, 5]
Subconscious Execution: Trusting your body to execute complex movements without overthinking every minor muscle twitch. [2, 4]

Shifting Between Effort and Release

Expert climbers constantly manage their attention to prevent burnout and anxiety: [6, 7, 10, 11]

Spotlight Attention (Assessment & Analysis): Consciously analyzing a crux move, looking for footholds, or managing risk.
Ambient Attention (Sensory & Flow): A relaxed, broad focus where you allow your instincts and muscle memory to take over while climbing through known territory.
Transitioning: Knowing when to pause, rest, and assess a situation versus when to commit and move continuously. [4, 6, 10, 12, 13]

How to Train Your Attention

To improve your mental game and performance on the rock, try incorporating these drills into your climbing sessions: [2, 14, 15]

Visualization & Route Reading: Before pulling onto the wall, memorize the route. Deliberately identify your handholds and footholds. [4, 16, 17, 18, 19]
Body Awareness Check-Ins: While climbing, ask yourself micro-questions: Am I pulling too much with my arms? Are my hips positioned correctly? Am I over-gripping? [12]
Breathing and Momentum: Maintain continuous, calm belly breathing and use momentum to keep moving and conserve energy. [10]
Committed Moves Drill: Choose an intimidating move and attempt it with a deliberate setup, visualization, and a simple mantra to eliminate mid-move hesitation. [20]

End Hesitation in Your Climbing: Committing Moves Drill, YouTube · Paradigm Climbing · 2026 M03 9 If you’d like to explore this further, let me know:

What specific climbing discipline are you interested in (e.g., bouldering, sport lead, or top rope)?
What is your current experience level?

I can share targeted drills and specific techniques tailored to your needs.

[1] https://medium.com [2] https://www.ukclimbing.com [3] https://prowesscoaching.co.uk [4] https://www.ukclimbing.com [5] https://www.powercompanyclimbing.com [6] https://www.maxclimbing.com [7] https://www.researchgate.net [8] https://www.linkedin.com [9] https://www.theclimbinghangar.com [10] https://www.climbing.com [11] https://altaclimbing.com [12] https://www.instagram.com [13] https://www.twostonesclimbing.com [14] https://www.researchgate.net [15] https://www.labyrinthclimbing.co.uk [16] https://pmc.ncbi.nlm.nih.gov [17] https://butorausa.com [18] https://www.safra.sg [19] https://blog.movementgyms.com [20] https://www.youtube.com

additional prompt “‘m interested to contrast with knowledge work, or self-development efforts and just how to intentionally live life. while i like the concept of the 4Cs (cal newport) i feel there is a link to knowledge work and how much time to devote to different aspects of life and software development skills growth. and when to keep persisting and when to save your energy and try another way on another day”:

The link between climbing and knowledge work—specifically software development—is found in the economy of attention. In both fields, you have a limited reservoir of “send power” (the ability to execute at your limit), and mismanaging it leads to falling off the wall or burning out on a codebase. Cal Newport’s 4 Disciplines of Execution (4DX)—the “4Cs” framework often associated with Deep Work—serves as the bridge to intentionality. [1, 2]

1. The Strategy of Persistence: Crux vs. Flow

In climbing, you don’t climb at 100% effort the whole way; you “over-grip” during the crux (the hardest move) and relax during the rest.

When to Persist: Persist when you are in a “deep work” window (Rule #1) and facing a specific, reachable technical hurdle. Like a climber who has already committed to a sequence, backing off mid-move is more exhausting than finishing it. In code, this is the final 20% of a refactor or a complex bug fix where you have the full mental context. [3, 4, 5]
When to Save Energy: If you are experiencing “attention residue”—feeling like your brain is still stuck on the last meeting or an unread Slack message—you are essentially “flash-pumped”. Save your energy when your “lead measure” (time spent in depth) is dropping. In climbing terms, if your form is failing and you’re just “slapping” at holds, it’s better to descend, rest, and try another day with a fresh perspective. [6, 7, 8, 9]

2. Time Devotion and the “Lead Measure”

Newport suggests that even elite performers can only sustain about 4 hours of deep work per day. [10, 11]

Software Skills Growth: Treat this as a “Wildly Important Goal” (WIG). Don’t try to “learn everything” every day. Devote your first 90–120 minutes of peak energy strictly to skill growth or complex architecture.
The Scoreboard: Track hours of deep work, not tasks completed. This mirrors a climber’s “training log” rather than their “summit list.” [1, 3, 8, 12]

3. Living Intentionally: Transitioning

Intentional living is the “Shutdown Ritual” of life. [13, 14]

Drain the Shallows: Aggressively limit “shallow work” (emails, administrative tasks) to specific batches. This is like the approach to a mountain: you don’t want to carry unnecessary gear (mental baggage) that doesn’t help you reach the top. [4, 15]
Embrace Boredom: Just as a climber must be comfortable standing on a small ledge for minutes to recover, you must train your mind to handle the absence of digital stimulation. This “productive meditation” builds the capacity to stay on the “wall” of a difficult problem longer. [4, 16, 17]

Summary of the “Knowledge Climbing” Framework

Concept [6, 8, 10, 13, 18]	Climbing Parallel	Knowledge Work / Dev Application
Deep Work	The “Send” Attempt	2–4 hours of focused coding/learning.
Lead Measure	Time on the wall	Tracking “Deep Work Hours” on a scoreboard.
Attention Residue	Being “Pumped”	The cognitive lag from checking Slack/Email.
Shutdown Ritual	Coming off the mountain	A hard mental break to allow for recovery.

Would you like to see a sample weekly schedule that balances software skill growth with professional output using these “climbing” mental models?

[1] https://medium.com [2] https://frankcjones.com [3] https://www.flashover.blog [4] https://paso.to [5] https://medium.com [6] https://medium.com [7] https://www.parkerklein.com [8] https://medium.com [9] https://medium.com [10] https://wellnessdivision.com [11] https://goalsandprogress.com [12] https://paso.to [13] https://www.youtube.com [14] https://www.jvandemo.com [15] https://towardsdatascience.com [16] https://dansilvestre.medium.com [17] https://medium.com [18] https://avthar.com

2026 Week 21 - AI agent flows

Sun, 24 May 2026 00:00:00 GMT

2026W21

TLDR;

Compare and contrast Matt Pocock AI workflow, particularly around the use of skills /grill-with-docs different types of job types of Human In Loop vs. Away From Keyboard with Allium language and approach.
Try Matt Pocock skills, compare the improve codebase architecture skill with modularity tools from - GitHub - vladikk/modularity: Claude Code plugin for designing modular systems from functional requirements and reviewing existing codebases for modularity problems with actionable improvement recommendations. Based on the Balanced Coupling model. · GitHub and GitHub - JuliusBrussee/caveman: 🪨 why use many token when few token do trick — Claude Code skill that cuts 65% of tokens by talking like caveman · GitHub.
Compare results with a varion Kit approach. Been a slow month on prof. dev but I’ve come back to catch up on AI agent flows and particularly about a good approach and scaffolding to achieve good outcomes for AI adoption. While tooling is necessary, I want practical heuristics that can be applied generally.

Matt Pocock has done a decent job explaining his flows and how he focuses on skills to define the job to be done / prd and then how to get the issues/stories into vertical slices/tracer bullet (for pragprog) to guide towards modules with deep interfaces (Jonh Ouster)

I’ve got a growing list of skills to go back to.

2026 Week 17 Thu

Thu, 23 Apr 2026 00:00:00 GMT

Thu 23-04-26 - 2026w17-4

Interesting Wardley maps claude skill

2026 Week 16

Tue, 14 Apr 2026 00:00:00 GMT

2026W16

I’m on holidays this week but hope to try gemma4 locally with - Running OpenCode with Gemma 4 26B on macOS via llama.cpp - fixing tool calling · GitHub Vladd, author of Balanced Coupling in Software Design has released skills I’m keen to try - See GitHub - vladikk/modularity: Claude Code plugin for designing modular systems from functional requirements and reviewing existing codebases for modularity problems with actionable improvement recommendations. Based on the Balanced Coupling model. · GitHub.

Keep a lid on your token spend with Caveman - GitHub - JuliusBrussee/caveman: 🪨 why use many token when few token do trick — Claude Code skill that cuts 65% of tokens by talking like caveman · GitHub

Maybe I’ll get time to try this also - GitHub - ultraworkers/claw-code: The repo is finally unlocked. enjoy the party! The fastest repo in history to surpass 100K stars ⭐. Join Discord: https://discord.gg/5TUQKqFWd Built in Rust using oh-my-codex. · GitHub

Try Spec Kit, need to try again.

Thu, 02 Apr 2026 00:00:00 GMT

2026W14-5

Bit under-whelming, vs. general vibe results. I probably need to give it a generally well known problem to remix and regurgitate.

2026 Week 14 - new category

Wed, 01 Apr 2026 00:00:00 GMT

2026W14

Simplest zig wasm

export fn add(a: i32, b: i32) i32 {
    return a + b;
}

zig build-exe add.zig -target wasm32-freestanding -fno-entry -rdynamic

<!doctype html>
<html>
  <body>
    <script>
      const request = new XMLHttpRequest()
      request.open('GET', 'add.wasm')
      request.responseType = 'arraybuffer'
      request.send()
      request.onload = function () {
        WebAssembly.instantiate(request.response, { env: {} }).then((result) => {
          const add = result.instance.exports.add
          console.log(add(3, 5)) // Outputs 8
        })
      }
    </script>
  </body>
</html>

run

python -m http.server

WASM - Simplest Thing That Works (STTWs)

Wed, 01 Apr 2026 00:00:00 GMT

2026W14-3

Simplest zig wasm

export fn add(a: i32, b: i32) i32 {
    return a + b;
}

zig build-exe add.zig -target wasm32-freestanding -fno-entry -rdynamic

<!doctype html>
<html>
  <body>
    <script>
      const request = new XMLHttpRequest()
      request.open('GET', 'add.wasm')
      request.responseType = 'arraybuffer'
      request.send()
      request.onload = function () {
        WebAssembly.instantiate(request.response, { env: {} }).then((result) => {
          const add = result.instance.exports.add
          console.log(add(3, 5)) // Outputs 8
        })
      }
    </script>
  </body>
</html>

run

python -m http.server

2026 Week 13 - Agent comparisons - AgentOS, BMAD, SpecKit

Mon, 23 Mar 2026 00:00:00 GMT

2026W13

tbh, the bottom up and faster tools of Spec Kit and Open Spec I’ll probably lean towards.

But I’ll still need to try out BMAD and AgentOS for 1st hand experience.