Stars
Terminal-Bench Science: Evaluating AI Agents on Complex Real-World Scientific Workflows in the Terminal
turns your codebase into an autoresearch loop — discovers what to measure, instruments the benchmark, then runs tree search with parallel subagents.
A set of delightful extensions for Pi
tmux sidebar for coding agents — Amp, Claude Code, Codex, OpenCode. Per-thread markers, local HTTP API, live session state.
🛰️ Track token usage across AI coding agents from your terminal. 🏅 Global leaderboard with quadrillions of tokens tracked.
Democratizing AI scientists with ToolUniverse
AlphaFast: ultra-high-throughput AlphaFold3 inference with MMSeqs2-GPU
A minimal PyTorch re-implementation of AlphaFold2's model & training
Persistent Context Across Sessions for Every Agent – Captures everything your agent does during sessions, compresses it with AI, and injects relevant context back into future sessions. Works with C…
"Deep Generative Modeling": Introductory Examples
Fast protein backbone flexibility prediction model
OVO, an open-source ecosystem for de novo protein design
Proxy that exposes Antigravity provided claude / gemini models, so we can use them in Claude Code and OpenClaw (Clawdbot)
DE-STRESS is a model evaluation pipeline that aims to make protein design more reliable and accessible.
FoldBench is a low-homology benchmark spanning proteins, nucleic acids, ligands, and six major interaction types, enabling assessments that were previously infeasible with task-specific datasets.
Code for ImmunoGeNN: Accelerating Early Immunogenicity Assessment for Generative Design of Biologics
Create minimal docker images from conda environments
TLimmuno2: predicting MHC class II antigen immunogenicity through transfer learning
An IPython/Jupyter widget for interactive molecular visualization, based on Molstar. This is the Jupyter widget version of nano-protein-viewer.