If you looked at my GitHub profile you'd see a graveyard: 11 repos, each with two or three weeks of commits, then silence. A CLI budgeting tool. A home inventory app. A static site generator nobody asked for. I'm 34, I write backend code for a living, and for years every side project has died the same death — usually around week three, when the boring parts start.
So when my wife came home from another trip to the rock shop with an orange stone she said was for creativity, my first instinct was to laugh. My second instinct was to open a spreadsheet.
What follows is the 8-week experiment I ran this spring, with receipts. Short version: the rock did nothing, and I shipped my first side project in 14 months. Both statements are true at the same time.
The setup
I bought three stones, because crystal folks say the sacral chakra — supposedly the creativity department — responds to orange stones:
- Carnelian, tumbled, $19
- Orange calcite, raw chunk, $12
- Sunstone, small pocket stone, $9
Total: $40. Yes, that is one mechanical keycap's worth of money. Moving on.
Protocol, because I don't know how to do things casually:
- Four weeks of baseline first (March): counted git commits to side-project repos only — work commits don't count. Also logged pomodoros (25-minute timer) spent on side projects after 8pm.
- Baseline came out to 2.3 commits/week and 2.5 pomodoros/week. One week was a literal zero. Eleven projects, zero shipped, fourteen months.
- Then eight weeks with the carnelian sitting on my desk by the keyboard. The rule: stone on the laptop lid, 25-minute timer, no phone until it rings. That is the whole ceremony. No meditation, no moonlight charging (my wife offered, I declined).
One honesty note before the results: I work from home, my evenings are mostly free, and I do not have kids. I am testing one variable on a life that already has spare hours in it. If your evenings look different, your mileage will differ.
What the numbers did
Weeks 1–2: 13.5 commits/week, 7 pomodoros/week. Before anyone concludes the carnelian works — that is textbook novelty effect. I get the same spike from a new notebook, a new pen, a new todo app. It is New Thing energy, not geology.
Weeks 3–4: 6 commits/week. The spike decayed right on schedule, same as every productivity toy I have ever bought, and this is exactly where my projects usually die. This time something stuck: by week 3 the stone-and-timer routine had stopped being a decision and become a reflex. I no longer spent 40 minutes negotiating with myself about whether tonight was a coding night. I just did the 25 minutes, and half the time one pomodoro turned into three.
Weeks 5–6: 8.5 commits/week — but my wife confiscated the sunstone for her own desk mid-stretch and I dipped to 4 commits one week. Coincidence, probably. n=1, remember. The calendar view was telling too: 63% of my side-project pomodoros happened Tuesday to Thursday. Mondays I was fried from work, Fridays I wanted nothing to do with a screen. Knowing that let me stop feeling guilty about two evenings a week, which weirdly made the other five easier.
Weeks 7–8: 7 commits/week, and the headline: I shipped. A 214-line Go CLI that converts markdown tables to CSV, with an actual README. First finished thing in 14 months. Nobody has starred it. My wife baked a cake anyway.
Final tally: 6.1 commits/week average across the 8 weeks vs 2.3 at baseline. Pomodoros 5.8/week vs 2.5. Cost per shipped project: $40, the cheapest intervention I have ever tested, and I include actual therapy in that comparison.
What I think actually happened
The carnelian did nothing. It is silicon dioxide; it emits motivation at exactly zero hertz. What it did do was turn work-on-the-side-project from a decision into a trigger, and decisions were the thing killing my projects. The nightly 9pm negotiation — do I code tonight or do I deserve rest — is one I lost about 70% of the time. The ritual deleted the negotiation. Behavior nerds call it an implementation intention. Habit people call it a cue. My wife calls it the orange rock system. Same machinery.
The placebo component is real and I am not even mad about it. If a $19 rock is what gets the laptop open at 8:40pm instead of the scroll, the mechanism being psychological is not a bug. It is the entire product.
If you want to steal the system
The stone is optional. The ritual is not. The parts that matter:
- Pick one physical object you only touch during side-project sessions — stone, keycap, stupid mug, whatever. It is a switch, not a decoration.
- Keep it on the desk, visible. Objects in drawers trigger nothing.
- One 25-minute timer. The contract is the 25 minutes, not the motivation. You are allowed to stop when it rings.
- Track commits, not vibes, for at least a month. Numbers dismantle self-deception faster than journaling ever will.
- Expect week-3 decay. It is coming for you too. The object is what carries you through it.
If you are curious about the framework I borrowed all this from: mid-experiment I went down the rabbit hole on chakra stuff, purely as research, obviously. The clearest thing I found on the creativity angle was this sacral chakra crystals guide — that is where the orange-stones idea came from, and yes, I rolled my eyes the whole time I read it. If your blocker is more terrified-to-publish than cannot-start, the writeup on solar plexus stones for confidence fits better — I suspect that describes half of us. New to the whole topic? The beginner's crystal healing guide is the least unhinged on-ramp I found, and the rest of SagStone's crystal blog reads like people who genuinely believe this stuff, which I consumed the way I watch cooking shows: for the craft, not the belief system.
Where it stands now
Four months after the experiment ended, the carnelian is still on my desk. It got promoted from experiment to infrastructure. The repo count is now 10 unfinished, 1 shipped, which sounds unimpressive until you have been the person with 11.
Last week my wife ordered lapis lazuli for my focus, which per this third eye chakra guide is apparently the stone for exactly that. I already built the spreadsheet. Report in eight weeks.
Top comments (0)