by DrugClaw
π AI Research Assistant for Accelerated Drug Discovery. π¦
# Add to your Claude Code skills
git clone https://github.com/DrugClaw/DrugClawLast scanned: 5/31/2026
{
"issues": [],
"status": "PASSED",
"scannedAt": "2026-05-31T07:46:16.215Z",
"npmAuditRan": true,
"pipAuditRan": true
}No comments yet. Be the first to share your thoughts!
DrugClaw is an AI Research Assistant for Accelerated Drug Discovery, implemented as a Rust multi-channel agent runtime. One agent core serves chat channels, the local Web UI, hooks, scheduled tasks, and domain skills without splitting the product into separate bots.
drugclaw setup, drugclaw doctor, and drugclaw start.This project is built on top of microclaw.
DrugClaw combines drug-discovery research workflows with a general-purpose agent runtime:
Current channel adapters include:
DrugClaw is already useful as:
The runtime is generic enough to automate other workflows, but the product direction is explicitly drug-discovery research acceleration.
DrugClaw is strong at:
DrugClaw is not:
When DrugClaw reports docking scores, QSAR predictions, ADMET heuristics, or affinity estimates, those outputs should be treated as prioritization signals only.
Below are live demonstrations of DrugClaw handling real tasks via Telegram.
Fetch a PDB structure, render it in rainbow coloring with PyMOL, and send the image.
Search PubMed for recent high-impact papers and provide structured summaries.
Visualize hydrogen bonds between a ligand and protein in PDB 3BIK.
Build a concise target dossier by combining UniProt, OpenTargets, Reactome, STRING, ClinVar, and known-drug evidence into one brief.
Query PubChem, ChEMBL, and BindingDB for a compound or target, normalize the returned activity records, and send back a ranked summary table.
Generate the search box, run docking, and return the top poses with a compact report.
curl -fsSL https://drugclaw.com/install.sh | bash
When Docker is installed and the daemon is reachable, the installer also tries
to build the default science sandbox image drugclaw-drug-sandbox:latest.
iwr https://drugclaw.com/install.ps1 -UseBasicParsing | iex
git clone https://github.com/DrugClaw/DrugClaw.git
cd drugclaw
cargo build
npm --prefix web install
npm --prefix web run build
./uninstall.sh
cp drugclaw.config.example.yaml drugclaw.config.yaml
drugclaw setup
drugclaw doctor
If the default sandbox image is already present locally, drugclaw setup
defaults the bash sandbox to enabled.
drugclaw start
By default the UI listens on http://127.0.0.1:10961.
A smallest practical config is usually Web-first, then add channels as needed.
llm_provider: "anthropic"
api_key: "replace-me"
model: ""
data_dir: "./drugclaw.data"
working_dir: "./tmp"
working_dir_isolation: "chat"
channels:
web:
enabled: true
telegram:
enabled: false
web_host: "127.0.0.1"
web_port: 10961
Recommended next steps:
channels:soul_path or add SOUL.mddrugclaw web password-generate for Web operator accessThe runtime does one thing consistently across channels:
The shared loop lives in src/agent_engine.rs. Channels are ingress and egress adapters, not separate agent implementations.
DrugClaw has two memory layers:
AGENTS.md plus chat-scoped files under runtime/groups/This lets the runtime keep durable context without forcing every instruction into a single prompt.
Bundled skills currently include:
bio-toolsbio-db-toolsbayesian-optimization-toolsomics-toolsgrn-toolstarget-intelligence-toolsvariant-analysis-toolspharma-db-toolschem-toolspharma-ml-toolsliterature-review-toolsmedical-data-toolsclinical-research-toolsmedical-qms-toolsstat-modeling-toolssurvival-analysis-toolsscientific-visualization-toolsscientific-workflow-toolsdocking-toolsBundled domain skills now cover:
See [docs/o