ideation, no reading, this is my book report

I've felt inspired by this vacation, and some of the inspiration-energy I'm experiencing is taking shape in thoughts of what games or "tools" (which tbh amount to games, just with marginally more relevancy to the real world).

ideas

anagram animations - I can totally do this! I'm good at anagrams! Just take a starter word, display it on a page nice and big, and then animate the individual letters (sliding, rolling, etc) into their positions in the new word. Bonus: "strange garnets" (almost-igrams) animations, which are basically anagrams but perhaps one of the letters must be deleted or created in the process.
It wouldn't be super hard to turn this into a game, where the player must guess the anagram or strange garnet, and perhaps it helps them overcome an obstacle in a manner appropriate to the word or phrase formed. For instance, starting with a tall-and-pointy-eared slender humanoid in an orange or black-and-white-stripes jumpsuit blocking your path, and the word "felon", you may type "no elf", and it disappears, unblocking your path. Perhaps more opportunity is opened for less prompting, if instead in that scenario, there were both dwarven and elven convicts, and you could type in the whole phrase -- with enough context clues like back-and-forth banter indicating one group is dwarves and the other elves, and that "[character] wish[es] they weren't there, then I could get through". Another way to solve that puzzle may be "forward, o dwarf!"to get one of the dwarves to take a step (out of the way). Maybe both are required.

lightsaber game; single click is slash, double-click is 360-slash. Audio is Owen Wilson saying wow, long audio is Owen Wilson saying wow, but paulstretched.

Audio time-compression software ideas. De-paulstretch -- take it down to its representative frequencies. But you'd end up with cacophony, right? Perhaps that's fine, given efficient-Big-Brother alternatives, but perhaps...
Choose the best-represented, and least-represented frequency ranges, and play those pairwise (a low range with a high range), segmenting frequency ranges by a given scale (max range width, adjusted for exponential scale) and "tolerance range" (if a set of frequencies is less than x (exponentially scaled) Hz apart, it counts as "in the same range" for purposes of grouping)
Or just capture entire segments that fit a certain amplitude threshold under a given filter (which may usually be specified to look for human voice, cat voice, and/or motor sounds -- may want to run tests to see what filter rules capture what audio, and what thresholds are appropriate)

Finish off the rpi audio greeter project

Finish off the rpi audio hub project