Mark Hesketh (@markahesketh) header
Mark Hesketh (@markahesketh) profile picture

Ruby on Rails, React, Hotwire etc. 🤓 Full Stack Software Engineer 🇬🇧 Northerner

Liverpool, UK
Joined April 2010
Photo by Mark Hesketh (@markahesketh)Photo by Mark Hesketh (@markahesketh)Photo by Mark Hesketh (@markahesketh)Photo by Mark Hesketh (@markahesketh)Photo by Mark Hesketh (@markahesketh)Photo by Mark Hesketh (@markahesketh)Photo by Mark Hesketh (@markahesketh)Photo by Mark Hesketh (@markahesketh)Photo by Mark Hesketh (@markahesketh)Photo by Mark Hesketh (@markahesketh)
Might have to start orbin'. My laptop is on fire 🔥
21
Started using pstack and I'm tokenmaxxing like its Christmas 2025 again
28
Still using Cursor's `thermo-nuclear-code-quality-review` skill regularly: github.com/cursor/plugins/bl…
44
Tests are still worth it, maybe more than ever. I use this rule to keep my agents from death spiralling:
1
1
48
Add to your AGENTS.md or whatever: Test each behaviour once, at the cheapest level that fails when it breaks. No duplicating coverage across levels. Don't mock the thing under test to force a test lower — real collaborators mean it belongs higher.
18
I regularly ask the agent "are you proud of this work?". It never is and does a better job.
im probably hallucinating this but telling astra to "think from first principles" is like +20 iq points hack
2
97
At work we enabled GitHub status notifications in Slack.. then disabled it hours later due to spam 🙃
1
66
Anecdotal.. but starting to see people posting about other topics rather than just AI. Nice.
47
Saw this and started building my own productivity app 🫠
The only project management you need until IPO.
1
2
120
2 Mostly Technical pods in 1 week. Feeling #blessed
52
Unironically gonna try this for a couple of days
despite my pith here, if you hate seeing "load bearing" and "it earns X...", asking a model to use ASD-STE100 is probably your best lever for now
94
Enjoying Pi. Took 10 mins to set up the stuff I liked from CC and now I can run both Claude and Codex in the same TUI/session
4
364
Planning an MMORPG with Godot and Elixir. Sol went architecture astronaut over it. 13 grilling sessions, just insane stuff. Luna on max seems to be doing a good job, staying pragmatic, good questions and more focused.
1
87
Eval'd GPT models building a React UI tonight. All test passed, looked similar etc. Sol/med, no subagents ~$2.50 Sol/med orchestrator, Luna/max implementors: ~$1.85 Luna/max orchestrator and implementors: ~$0.25 Gonna use Luna a lot more
77
Frontend devs don't read the AI's backend code. Backend devs don't read the AI's frontend code.
71
This reduced my initial token use from 33k to 19k. Cool cool cool
Here's a step-by-step process to kill all the bloat from your Claude Code system prompt: 1. Run a proxy so you can see exactly what gets sent to Claude Code (included in the article) 2. "Fuck, there is so much cruft in there" 3. Use my settings.json to kill all the bloat Down to a clean 13K tokens to start each session with. Nice. Full process here: aihero.dev/how-to-kill-the-b…
125
Getting to write Rails professionally is a joy
i feel sad for anyone who doesn’t appreciate ruby on rails
3
152
Just realised - instead of bloating context with rules that get ignored, Claude can write custom linter rules that enforce my preferences. e.g. a Rubocop cop requiring resourceful controllers (index, new, create, etc only)
1
70
TIL you can use !`command` in Claude skills and it'll inject the output into the context automatically. Its done before its sent, so Claude gets the data not the command.
2
65