Your codebase deserves memory.
Check hoy my #ClaudeCode / #Codex #plugin surfaces the relevant past learning automatically on every prompt, captures new ones on its own and keeps them fresh. Local, no per-prompt cost.
Just updated it to v0.3.0: sharper recall, get repo link ⬇️
Lore is ambient memory for Claude Code & Codex: it recalls the relevant past learning automatically on your next prompt (no slash command), captures new ones on its own, and keeps them fresh. Local, zero recurring cost.
github.com/aoc81/lore
Every codebase has lore... the bug with the non-obvious cause, the convention nobody wrote down, the footgun that ate an afternoon. It lives in people's heads and old chat threads. So I built Lore: a CC/Codex plugin that writes it down where the work happens. Why I made it 👇
It runs on both #ClaudeCode and OpenAI #Codex — same hook contract, same scripts, same store. macOS / Linux / Windows, stdlib Python, no build step, no dependencies. MIT licensed.
TL;DR: I wanted my codebase to have a memory, so the same lesson is never learned twice. Capture once, recall forever.
github.com/aoc81/lore
Most work also accrues debt: each change makes the next one a little harder. I wanted the opposite gradient. Capture what a task taught you, feed it forward, and the next task starts easier. Compound interest, but for hindsight.
This project started with a name.
A Russian wartime unit called itself #Española.
As a Spaniard working in #OSINT, I couldn’t just scroll past that.
Ten months later, it had become a book. Get a free copy in spainrus.org
I used AI while writing this book.
Not as a fact machine. Not as a shortcut. It helped me structure, draft, cut repetition, test chapters, and pressure weak passages until they either improved or collapsed.
That was the useful part. spainrus.org/read/a-note-on-…
The interesting part was not just using AI.
It was building a workflow around it.
Notes. Fragments. Source extracts. Drafts.
Then protocols.
Then revision.
Then editorial pressure.
That whole process is public here:
spainrus.org/codex
Good example about how #AI can help you to gather intelligence. Listing Chinese air, naval, and land military bases #OSINT poe.com/s/bAjFxVSuopm9baDiWZ…
#PrivateGPT
Absolutely necessary in all contexts but specially when handling sensitive information. Thanks for it @ivanmartit github.com/imartinez/private…
This is a great resource to support Osint investigations, it will be also very useful to debunk fake images.
With our “About this Image” tool in Search, you’ll be able to see important context, like where else an image has been seen online, or when and where similar images have appeared. Later this year, you’ll be able to use it in Lens in the Google app or Chrome. #GoogleIO