litman

Your papers, finally in order.

Every paper stays as a plain file on your own disk — and the AI agent you already use files, finds and links them for you.

macOS & Linux
$ curl -LsSf https://raw.githubusercontent.com/wqx1999/litman/main/install.sh | sh
Windows
> irm https://raw.githubusercontent.com/wqx1999/litman/main/install.ps1 | iex
Scroll

The problem

Where does it all go?

Every paper you meant to understand ends up as one more cryptic PDF in a pile.

~/Downloads
  • PDFpaper (3).pdfMar 4 · 8.1 MB
  • PDF1706.03762v5.pdfMar 4 · 2.2 MB
  • PDFnihms-1832902.pdfFeb 19 · 5.7 MB
  • PDF41586-2023-06221-x.pdfFeb 11 · 12 MB
  • PDF1706.03762 (2).pdfJan 28 · 2.2 MB
  • PDFUntitled.pdfJan 22 · 940 KB
  • PDFfinal_v2_REAL.pdfJan 9 · 3.8 MB
  • PDFdownload.pdfDec 30 · 6.4 MB
Which folder did I even put it in?
Wait — did I read this one already?
I had thoughts on this. They're gone.

The engine · your agent

Just ask, and it's filed.

Ask the AI agent you already use, in plain words. It runs the commands — filing, finding and linking papers for you.

claude · litman
Add the paper I just downloaded — it's about peptide binders.
lit add ~/Downloads/2310.19325v2.pdf lit modify 2024_Chen --add-tag topics=peptide-design --set priority=A
Filed under peptide-design, priority A, in papers/2024_Chen_Diffusion-Peptide-Binders/. It cites two papers already in your vault — link them?
Which paper did I read on diffusion models for peptide design?
lit list --topic peptide-design --status read --sort recent
That's Chen 2024 — you read it three weeks ago. Your notes.md flags the sampler as the weak point. Open it?
Clone this paper's code and check the training script runs.
lit code add 2024_Chen --repo github.com/chen-lab/diff-peptide cd codes/2024_Chen && python train.py --dry-run
Cloned into codes/2024_Chen/. train.py needs torch 2.1 and a config they didn't ship — I drafted one and noted it in discussion.md.
What else in my library connects to this one?
lit list --method diffusion --topic peptide-design lit relations 2024_Chen --add related=Watson2023,Ingraham2023
Three neighbours already in your vault: Watson 2023 · Ingraham 2023 · Badrinarayanan 2024. I linked the closest two as related.

The window · for you

Sit down and read.

The agent keeps everything filed and linked. The app is that same library, made for reading — where you highlight, annotate and note.

Filter by status, priority, topic — the same library the agent files into. Read it in place. Highlight and annotate; the marks stay with the paper. Notes are Markdown on disk. The agent reads them; so does any editor you like.

Filtering the litman library by status, priority and topic
Reading a PDF inside litman
Markdown notes inside litman

Start with one paper.

macOS & Linux
$ curl -LsSf https://raw.githubusercontent.com/wqx1999/litman/main/install.sh | sh
Windows
> irm https://raw.githubusercontent.com/wqx1999/litman/main/install.ps1 | iex