Engineers are used to reactive debugging: an alert fires, a test fails, a customer files a ticket, and we drop everything. The harder problem is the slow drift — the week that quietly slides into Friday carrying seven half-finished threads, two missed commitments, and one conversation you should have had on Monday. A daily three-card check-in borrows the same idea as a morning stand-up or a test-driven development cycle: small, cheap, repeatable checks that surface drift before it compounds.
This article walks through how to run a three-card pull every morning, what each position should represent, and how to convert the result into concrete actions inside your calendar and code-review queue. The framing is different from a "big decision" spread — this is a maintenance ritual, not an oracle. Where you want a one-time deep walkthrough on the yes/no pattern that powers the daily pull, the Free Yes or No Tarot Reading Online: Three Cards guide covers the mechanics in detail.
Why a daily ritual beats a quarterly retreat
Most engineers I have worked with resist "reflection" exercises because they feel vague. Reflection only pays off when it produces an artifact you can act on by lunch. Three cards is small enough to fit into the time between brewing coffee and opening your inbox, but structured enough that you cannot cheat it with vague affirmations.
Think of it as a lightweight chaos engineering experiment on your own week: you inject a tiny probe on Monday morning and observe what surfaces. The probe is not the cards — the probe is the question you bring to them. The cards force you to commit to a question, and the act of naming the question is where most of the value already lives.
The other reason a daily ritual beats a quarterly review is signal density. A once-a-quarter tarot session tries to summarize three months of context into a single reading; you spend the first twenty minutes just remembering what you were worried about. A daily pull runs in fresh context, so each position can be sharp.
The three positions and what they actually mean
A three-card spread can mean almost anything, which is the point and the trap. The trap is that you reach for grand interpretation: the Tower means my project will collapse. The point is that you define the positions yourself and hold them loosely. Here is a working definition I have used for two years.
Card 1 — Today's blocker. Draw one card for the single thing most likely to slow you down today. Not the biggest project, the blocker. The thing that, if you do not name it, will ambush you at 2 p.m.
Card 2 — Today's leverage. Draw one card for the one action that, if you actually do it, will unblock two other things. This is your Pareto move — the 20% that earns the 80%.
Card 3 — Tonight's honest answer. Draw one card for the question you are avoiding. The thing you have been dancing around for three days and need to write down, even if the answer is "I don't know yet."
The three positions together form a tight loop: name the obstacle, find the leverage, confront the avoidance. Run the loop daily and the week debugs itself.
A worked example: a senior engineer's Monday
Let me walk through a real-shape example, lightly anonymized. Monday morning, 8:40 a.m., before stand-up.
The blocker card is the Five of Swords. Translation in this context: I am going to win an argument today and lose a relationship. The concrete read: there is a thread on the design doc where two engineers are arguing about API shape, and I have been waiting for someone else to resolve it. If I "win" by pulling rank, I burn trust I will need on Thursday's cross-team review. Action: instead of replying in the thread, I DM one of the engineers and offer a 15-minute call to align before the meeting. The argument becomes a conversation.
The leverage card is the Three of Pentacles. Read: collaboration and craft. The concrete read: my highest-leverage move today is the code review I have been postponing on the payments service. It is blocking two other PRs and one onboarding ticket. Finishing the review unblocks three items and pays a colleague back for covering my on-call last week. I block 45 minutes after stand-up specifically for it.
The honest-answer card is the Moon. Read: something is unclear and I am pretending it is not. The concrete read: I have been telling myself the migration plan is "mostly done" but I have not actually opened the runbook since Thursday. I write a single line in my notes: "Migration runbook: status unknown, 30 minutes blocked at 4 p.m. to read it cold." I do not promise to fix it; I only promise to read it.
By 9 a.m. I have a Monday that is already shaped. None of this required the cards to be "right." The cards were the prompt; the engineering judgment was the work.
A workflow you can run in seven minutes
A ritual only sticks if it is faster than the resistance to doing it. Here is the workflow I use, with timings.
- Two minutes — frame the day. Write the date and one sentence on what today is for. Pull the three cards from a physical deck or a digital equivalent.
- Two minutes — name the blocker. Force one sentence per card, written, not thought. The sentence must name a specific artifact: a thread, a PR, a doc, a person.
- One minute — pick the leverage. Which of the three sentences has the most downstream dependencies? Block calendar time for it before you open Slack.
- Two minutes — schedule the honest answer. Put the avoidance item in the calendar with a verb and a duration. "Read migration runbook cold, 30 min" beats "look at migration stuff."
If a card genuinely does not spark anything, write the card name and a question mark, then move on. Silence is data; do not invent meaning. The point is the artifact, not the symbolism.
The workflow is deliberately short because the value compounds through repetition, not depth. A seven-minute daily ritual beats a forty-five-minute weekly one. The same principle applies to writing tests: a thousand fast unit tests catch more bugs than ten perfect integration tests, because the thousand get run.
Common pitfalls and how to steer around them
Treating the cards as predictions. The moment you start saying "the Tower says my deploy will fail," you have left engineering and entered superstition. The cards are a prompt for your own pattern recognition. If the card is "wrong," that is interesting data about your own framing, not evidence the system is broken. Hold the meaning lightly; hold the action tightly.
Vague artifacts. "Be more careful with code reviews" is not an artifact. "Review PR #4821 by 11 a.m." is. If your three sentences do not contain a noun you can put in a calendar, redo them.
Skipping the honest-answer card. This is the one people skip because it is uncomfortable. It is also the one that, six months in, you will not skip anymore, because you have learned it is the only place the actual risk surfaces. On the days you cannot name anything, the honest-answer card becomes "what am I pretending is fine?" — which is usually the right question.
Burning out on the ritual. If you miss a day, do not "make it up" with a five-card spread. The system fails when it becomes a chore. On heavy weeks, run a single-card version: just Card 3, the honest answer. Even that keeps the muscle alive.
Letting it become theater. If your team is interested, share the artifact (your three sentences) in stand-up, not the cards. The team does not need to know you pulled the Moon. They need to know you blocked 30 minutes for the migration runbook. The ritual is personal; the output is collaborative.
When a daily check-in is not the right tool
A daily ritual is wrong when you are in genuine crisis mode and the day needs a full incident response process, not a personal check-in. It is also wrong when you are using it to avoid a decision that actually needs a meeting, a design review, or a hard conversation. The cards can surface the avoidance, but they cannot resolve it for you.
If two days in a row produce the same blocker card, that is not the cards being repetitive — that is your week telling you something you have not heard yet. Stop pulling and go deal with the underlying item directly.
Finally, a daily check-in is not a substitute for therapy, mentorship, or a real retrospective. It is a maintenance habit. Treat it like a linter: useful, cheap, and not where the real thinking happens.
Frequently asked questions
How long does it take before the daily check-in actually helps?
Most people notice a difference in the second week. The first week is just learning to ask the questions in a useful way. By week three, the blocker/leverage/honest-answer pattern starts surfacing things you would otherwise have missed. The compounding effect is similar to writing a daily journal: the first entries are awkward, the thirtieth is invaluable.
What if I do not have a tarot deck?
You do not need one. A deck of playing cards works — just assign suits to the three positions (spades for blockers, diamonds for leverage, clubs for honest answers) and treat face cards as intensified versions. The point is the prompt, not the iconography. Digital tools are fine too; the only requirement is that the shuffle feels random enough that you cannot predict the draw.
Is this just journaling with extra steps?
Partly, yes. The cards are a randomized prompt generator that defeats your brain's tendency to reach for the same three comfortable answers. If you already have a journaling practice that produces the same artifacts (a named blocker, a leveraged action, an honest avoidance), you do not need the cards. The cards are for engineers whose reflective practice has gone stale and needs a small shake.
Can I share the results with my team?
Share the artifacts, not the cards. "I am reviewing PR #4821 by 11 a.m., then I am reading the migration runbook cold at 4 p.m." is useful stand-up information. "I pulled the Moon and it means my migration is unclear" is theater. If your team wants to adopt the practice, keep the cards private and standardize on the three-sentence output.
This article was drafted with AI assistance and reviewed for technical accuracy before publishing.
Top comments (0)