PyAutoMemory is the Memory of the PyAutoScientist — the long-term memory
of the organism: what it has learned, distilled into cross-linked LLM wikis —
literature summaries, scientific concepts, and the citation metadata to
verify them. Memory holds what the science says; what the organism did
lives in the Mind. Source PDFs live off-repo; what's here is the durable
knowledge. Start at index.md.
See the PyAutoMemory Dashboard
for managing all of it from one page: the reading
queue, the paper sections still needing a canonical citation key, and each
sub-wiki's maturity — every work queue carrying a one-tap 📋 button that
copies a paste-ready Claude prompt (file the next paper, resolve keys,
upgrade a stub, or /memory <domain> to recall what's known).
🧠 166 pages · 69 drafted · 40% of 645 paper sections cite a resolved key · 230 papers queued · dashboard →
Self-contained, shared schema:
| Wiki | Covers |
|---|---|
wiki/lensing/ |
strong gravitational lensing (the primary wiki) |
wiki/smbh/ |
supermassive black holes, binaries, recoil, GW background |
wiki/cti/ |
charge transfer inefficiency, Euclid VIS calibration |
wiki/methods/ |
Bayesian inference, samplers, deep learning, simulations |
wiki/galaxies/ |
galaxy formation and evolution |
bibliography/ holds the canonical BibTeX
metadata every wiki cites against; reading-queue.md
is what's waiting to be read and filed. New knowledge updates the metadata
and the claim support together, then passes make validate (CI-enforced on
every push).
The wiki schema is defined in
wiki/CLAUDE.md and inherited by every
sub-wiki. How agents should read this repo: AGENTS.md. The
organism this repo is the Memory of:
PyAutoBrain/ORGANISM.md,
documented in full at https://pyautoscientist.readthedocs.io.
Licence: structure and tooling MIT; wiki content CC BY 4.0 — see LICENSE.