Safely discover and use Therapeutics Data Commons datasets, splits, evaluators, benchmark groups, and molecular oracles via the PyTDC package.
A reference skill that guides Claude through scvi-tools deep generative models for batch correction, multimodal integration, and probabilistic differential expression in single-cell omics.
An installer skill that wires the NanoClaw container agent to local models served by the Atomic Chat desktop app via MCP.
A skill for ingesting VCF/BCF variant data into TileDB sparse arrays and querying or exporting it by sample and genomic region.
Wires an Emacs channel into NanoClaw so you can chat with your agent from a buffer or straight inside org-mode.
Sets up a persistent, self-maintaining wiki knowledge base in a NanoClaw group using Karpathy's LLM Wiki pattern.
Scans protein sequences for N-/O-glycosylation sites and guides glycan engineering for therapeutic antibodies and vaccine antigens.
A routing skill that guides installing, configuring, extending, and embedding Pi — the minimal terminal coding harness.
Installs and wires a native DeltaChat (email-based, end-to-end encrypted) messaging channel into a NanoClaw deployment.
Wires a NanoClaw agent into Linear issue comment threads so it answers every comment automatically.
Wires a Matrix channel into a NanoClaw project: copies the adapter, registers it, pins the dependency, and configures bot credentials.
Guides Claude to infer transcription factor–target gene networks from expression data using GRNBoost2/GENIE3.
Extracts Feishu/Lark docs, wikis, sheets, and Minutes transcripts into faithful local Markdown without letting the model rewrite the body.
Moves Claude Code's tool-locked personal memory into shared reference docs so Codex and other AI CLIs read the same user profile.
Turns markdown into production-grade Word (.docx) files via OpenXML, with battle-tested formal-document typography rules and a mandatory visual verification chain.
A maintainer-grade PR review workflow that judges contributor PRs by the real merge result against the current base branch, not a stale web diff.
Sets up isolated Claude Code CLI profiles so you can run Kimi, MiniMax, GLM, DeepSeek and Anthropic in separate terminal windows at once.
Pull the machine-readable Bigdata.com (RavenPack) substrate — financials, estimates, prices, daily sentiment, annotated chunks — through the official SDK and /v1 REST endpoints with built-in cost guards.
Diagnoses and fixes Tailscale conflicts with proxy/VPN tools (Shadowrocket, Clash, Surge, OrbStack/Docker) on macOS when SSH, curl, browsers or Docker mysteriously hang.
Fixes speech-to-text errors in two stages — a free dictionary pass plus Claude's native AI correction — and learns from every fix so the same errors auto-correct next time.