frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

Show HN: Animated step-by-step hints for any cryptic crossword clue

https://crypticinkling.co.uk
1•hiroprotag•42m ago
Cryptic crosswords are fun AND hard, and I've recently been introducing easier ones to my 11y/o daughter. It made me realise that when you're stuck, all you can do is hope you somehow come up with the way to crack it, or look at the answer and have all the satisfaction of figuring it out yourself removed!

I also realised that when I was explaining to her how clues worked, I would scribble all manner of letters, arrows doodles to make sense of how the words of the clue could be mangled and eventually produce the answer.

So I made a thing that would replicate the pencil and paper solving process, and create stepped visual animations that would enable you to get just enough info to make solving it within your reach, so you could feel like you did enough of the working out to get that "Aha!" feeling, without just getting straight to the answer.

You can paste in any cryptic clue with its length, e.g. "Around Los Angeles, coatless dudes make AI model, say (6)", and the animation shows you how to solve it, one step at a time. It's split into 3 or 4 sections, which most cryptic clues fit into:

1. Which end of the clue is the definition 2. What type of clue is it (there are several conventions in the cryptic world, a clue can be 1 of these, or a combination of several) 3. What are the pieces used to build the answer 4. How do the pieces fit together to create the answer

You can stop as soon as you have enough info to get you back to solving it yourself.

It was easy to get a smart AI to solve the clue - but it was trickier to get it to be reliable and correct. LLMs can be very good at writing a confident, plausible breakdown for a completely wrong answer! So I needed to use a multi-pass approach:

1. A small, fast model checks that the input looks like a cryptic clue. 2. A more capable model solves it and writes a step-by-step breakdown. 3. Another independent model solves it "blind", without seeing the first attempt or its answer. 4. A "judge" then checks that the two solvers agree on the answer and method, and that the letters in the breakdown actually add up to the answer.

If that fails, it tells you it’s unsure instead of bluffing, and if it’s only fairly sure that the answer and "how to" is the correct one, it says so.

The breakdown of how the clue is solved is then drawn as an animation, step by step, with letters and meanings scribbled and manipulated the way you'd do if you solving it yourself on paper.

Depending on the complexity, a clue usually takes between about 10 to 30 seconds to be solved. There's a cache, so clues that have already been solved once can be returned without going through the AI process again. It can still get things wrong, especially on loose clues or ones with some flexibility in how they can be interpreted, but generally it performs pretty well.

Your first clue solve is free, and an account gets you two more. There are paid plans to cover the AI costs involved for multiple solves as there is a cost to it.

There's also some good resources for learning cryptic crosswords basics - and a library of clues with hints to practice with. You can also write and share your own cryptic clues - along with the hint animation generated for them https://crypticinkling.co.uk/create

I’d love feedback from any cryptic crossword ninjas or beginners alike.

PS here's the breakdown for the clue at the top of this post: "Around Los Angeles, coatless dudes make AI model, say (6)" - how many hints do you need to crack it? https://crypticinkling.co.uk/clues/1hi0s2x

History of Spreadsheets (2026): VisiCalc to Taskade Genesis

https://www.taskade.com/blog/spreadsheet-history
1•mpweiher•1m ago•0 comments

Armory Castle Exploits BeyondTrust

https://cloud.google.com/blog/topics/threat-intelligence/armory-castle-exploits-beyondtrust
1•torutofu•3m ago•0 comments

Better Algorithms for Benign Bandits (2011) [pdf]

https://www.jmlr.org/papers/volume12/hazan11a/hazan11a.pdf
1•Eridanus2•5m ago•0 comments

SOPS: Secrets OPerationS Simple and Flexible Tool for Managing Secrets

https://getsops.io/
2•Bluestein•5m ago•0 comments

Evidence Before Trust

https://www.crunchbase.com/organization/eels-d60e
1•chuansoo•6m ago•0 comments

Natural screen reader voices on Linux

https://codemacabre.com/notes/natural-screen-reader-voices/
1•edent•7m ago•0 comments

Domain-Driven Design (DDD) – The Complete Guide – Visually Explained

https://www.youtube.com/watch?v=0c2fj6z6Zso
1•birdculture•8m ago•0 comments

Hyperoptimized Engineer 2026 [video]

https://www.youtube.com/watch?v=MhEL9cvLxQE
1•modinfo•9m ago•0 comments

Computer Ethics

https://en.wikipedia.org/wiki/Computer_ethics
1•DanielVZ•10m ago•0 comments

Ask HN: How much are developers trusting LLMs to write prod code?

1•devtor•11m ago•0 comments

LA restaurant group used $60k Dodgers bet to hedge playoff risk, Kalshi says

https://www.reuters.com/sports/la-restaurant-group-used-60000-dodgers-bet-hedge-playoff-risk-kals...
2•simonpure•18m ago•0 comments

Show HN: A daily deduction puzzle where one hex on a nautical chart is a lie

https://cartographerslie.com
1•kevmarsh91•21m ago•0 comments

Molecular 'scissors' open path to new therapies

https://newsroom.uw.edu/news-releases/molecular-scissors-open-path-to-new-therapies
1•geox•23m ago•0 comments

Make System One Models Fight

https://system-one-arcade.fly.dev/
1•tarasyarema•23m ago•1 comments

Strip Digital Signatures from PDFs Before Feeding Them to AI (LinuxGuide)

https://medium.com/@GiovambattistaVieri/strip-digital-signatures-from-pdfs-before-feeding-them-to...
1•gvieri•25m ago•0 comments

Super Intelligence

https://www.si.gov/
34•wiradikusuma•28m ago•48 comments

PrivateNotes: Encrypted notes, tasks, files, passwords, journal and bookmarks

https://github.com/LifetimeLabsDev/PrivacyNotes.app
1•akyuu•29m ago•0 comments

Emmanuel Maignan's gnomonic catoptric astrolabe at Palazzo Spada (2024) [video]

https://www.youtube.com/watch?v=gPQNDe8YNvo
1•zeristor•31m ago•1 comments

Regular Expressions: Now You Have Two Problems (2008)

https://blog.codinghorror.com/regular-expressions-now-you-have-two-problems/
1•simonebrunozzi•33m ago•0 comments

Why some people say the 2-day weekend is due for a makeover

https://www.cbc.ca/lite/story/9.7382232
2•colinprince•34m ago•0 comments

The Making of Grand Theft Auto VI

https://www.bloomberg.com/features/2026-grand-theft-auto-6-making-of/
2•mgl•34m ago•1 comments

Show HN: Animated step-by-step hints for any cryptic crossword clue

https://crypticinkling.co.uk
1•hiroprotag•42m ago•0 comments

The orca brothers preying on great white sharks

https://www.bbc.com/future/article/20261015-sharks-washed-up-one-after-another-after-another-why-...
1•neom•44m ago•0 comments

The Ugly Dell That Became My Dream Retro Gaming PC (2025)

https://www.jamesfmackenzie.com/2025/12/28/the-ugly-dell-optiplex-that-became-my-dream-retro-gami...
1•naves•44m ago•0 comments

What happens when AI can read our minds?

https://www.ft.com/content/45ea3dd0-080d-49a2-86e9-df0148929bf4
1•bookofjoe•46m ago•1 comments

Microplastics React with Water, a New Study Says. That's a Problem for Biology

https://www.wired.com/story/microplastics-react-with-water-new-study-says-thats-a-problem-for-bio...
2•beardyw•46m ago•0 comments

Show HN: Which Mac runs which LLM, and how fast

https://macyou.co/mac-llm-calculator
3•bagdaerdev•48m ago•0 comments

Monazite ore holds the key to powerful motor magnets–if we can handle it

https://spectrum.ieee.org/monazite-rare-earth
1•thread_id•48m ago•0 comments

Show HN: RAE FAANG – catalog and CLI for building products with coding agents

https://github.com/EnzoVezzaro/rea-faang
1•undergroundengi•50m ago•0 comments

Show HN: Offscreen - Give your agents a Mac screen of their own

https://github.com/markphelps/offscreen
1•bullcitydev•51m ago•0 comments