Skip to content
Back to search

DrugClaw

🦀 Agentic RAG for drug intelligence · 57 skills · 15 task categories · DTI · ADR · DDI · PGx · Repurposing · Powered by LangGraph

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today

Check it deeper on your machine

Its scripts were only partly scanned.

As a Claude Code plugin, install it and then ask Claude Code:

/plugin marketplace add keithmackay/mackayi/plugin install scanskillsafety@mackayiDo a deep scan of https://github.com/QSong-github/DrugClaw with scanskillsafety

Or from a terminal, clone it and run it:

git clone https://github.com/keithmackay/scanskillsafety-skill ~/.claude/skills/scanskillsafetynode ~/.claude/skills/scanskillsafety/cli.cjs --deep 'https://github.com/QSong-github/DrugClaw'

It checks the exact commit: every file, its packages against OSV (api.osv.dev), how it compares with this page's scan (findsafeskills.com receives the repository's address), and, if you agree, Claude reading every file with your own Claude subscription. It never runs the code.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/qsong-github--drugclaw/badge.svg)](/listings/qsong-github--drugclaw)

HTML

<a href="/listings/qsong-github--drugclaw"><img src="/api/listings/qsong-github--drugclaw/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

pip install -e .

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 170 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. <p align="center">
  2. <img src="./support/DrugClaw_Logo.png" alt="DrugClaw Logo" width="520" />
  3. </p>
  4. <p align="center">
  5. <strong>Agentic RAG for Drug Knowledge Retrieval, Reasoning, and Evidence Synthesis</strong>
  6. </p>
  7. <p align="center">
  8. <a href="https://arxiv.org/abs/2606.01434">📄 Paper (arXiv)</a>
  9. ·
  10. <a href="./README_CN.md">中文文档 / Chinese Version</a>
  11. </p>
  12. <p align="center">
  13. <em>Resource bundle will be released upon acceptance.</em>
  14. </p>
  15. <p align="center">
  16. <a href="https://arxiv.org/abs/2606.01434"><img alt="arXiv" src="https://img.shields.io/badge/arXiv-2606.01434-b31b1b"></a>
  17. <img alt="Domain" src="https://img.shields.io/badge/Domain-Drug%20Intelligence-1f6feb">
  18. <img alt="Registry" src="https://img.shields.io/badge/Registry-Source%20of%20Truth-0a7f5a">
  19. <img alt="Skills" src="https://img.shields.io/badge/Skills-Registry%20Driven-f59e0b">
  20. <img alt="Modes" src="https://img.shields.io/badge/Modes-GRAPH%20%7C%20SIMPLE%20%7C%20WEB__ONLY-7c3aed">
  21. </p>
  22. DrugClaw is a CLI and agent runtime for drug-focused questions. It prioritizes evidence-grounded retrieval, source attribution, and traceability over answers that only sound fluent.
  23. ## Why DrugClaw
  24. - It is not a general chat assistant. It is built for real drug questions such as targets, indications, repurposing, safety, DDIs, PGx, and labeling.
  25. - It aims to return structured evidence and source-grounded conclusions, not just a polished summary.
  26. - It supports both a lightweight minimal mode and a deeper full local-resource mode.
  27. ## Get Started in 5 Minutes
  28. If this is your first time using DrugClaw, follow the steps below and you should be able to run a real query quickly.
  29. ### 1. Clone the repository
  30. ```bash
  31. git clone https://anonymous.4open.science/r/DrugClaw-01A3
  32. cd DrugClaw
  33. ```
  34. ### 2. Install dependencies
  35. ```bash
  36. pip install -e .
  37. ```
  38. ### 3. Create `api_keys.json`
  39. DrugClaw reads `api_keys.json` from the repository root by default. If you do not want to pass `--key-file`, put the config file at the repository root.
  40. Create the file and provide at least these fields:
  41. ```json
  42. {
  43. "api_key": "<your-api-key>",
  44. "base_url": "<your-base-url>",
  45. "model": "gpt-5-mini"
  46. }
  47. ```
  48. If your config file lives somewhere else, pass it explicitly with `--key-file <path>`.
  49. ### 4. Verify the environment
  50. ```bash
  51. python -m drugclaw doctor
  52. ```
  53. If the setup is valid, you should see something like:
  54. ```text
  55. Doctor result: setup looks usable.
  56. ```
  57. ### 5. Run your first query
  58. ```bash
  59. python -m drugclaw run --query "What are the known drug targets of imatinib?"
  60. ```
  61. If this works, you already have the minimal usable setup running.
  62. ## What DrugClaw Is Good At
  63. - Drug targets and mechanisms
  64. - Indications and repurposing evidence
  65. - Safety risks and serious adverse reactions
  66. - Drug-drug interactions
  67. - Pharmacogenomics
  68. - Labeling and prescribing information
  69. ## Common Commands
  70. ```bash
  71. python -m drugclaw run --query "What pharmacogenomic factors affect clopidogrel efficacy and safety?"
  72. python -m drugclaw run --query "What are the clinically important drug-drug interactions of warfarin?"
  73. python -m drugclaw list
  74. ```
  75. ## Two Modes
  76. ### Minimal Mode
  77. Minimal mode is the default.
  78. The repository only tracks a minimal `resources_metadata/` subtree. That is enough for the CLI, basic queries, and the default test suite. For most new users, this is the right place to start.
  79. Use minimal mode if you want to:
  80. - try DrugClaw quickly
  81. - run the default queries and tests
  82. - avoid downloading large local resource bundles upfront
  83. ### Full Mode
  84. If you need deeper and broader local evidence coverage, the full `resources_metadata_full.tar.gz` bundle will be released upon paper acceptance. Once available, extract it at the repository root:
  85. ```bash
  86. tar -xzf resources_metadata_full.tar.gz
  87. ```
  88. Recommended flow (once the bundle is released):
  89. 1. Download `resources_metadata_full.tar.gz`
  90. 2. Extract it at the repository root
  91. 3. Run `python -m drugclaw doctor` again
  92. This is not just a generic data add-on. It expands the same `resources_metadata/` tree in place, enables more `LOCAL_FILE` resources, and supports deeper local evidence retrieval.
  93. Use full mode if you want to:
  94. - increase local resource coverage
  95. - enable more local-data-backed resources
  96. - run deeper resource-level validation or analysis
  97. ## What To Do Next
  98. - Show available resources and recommended entry points:
  99. ```bash
  100. python -m drugclaw list
  101. ```
  102. - Re-check your environment:
  103. ```bash
  104. python -m drugclaw doctor
  105. ```
  106. - Explore the built-in demo flow:
  107. ```bash
  108. python -m drugclaw demo
  109. ```
  110. # Citation
  111. If you find this project is useful for your research, please cite:
  112. ```
  113. Wang Q, Li B, Liang J, Shi D, Zhang B, Song Q. DrugClaw and DrugAudit: A primary-source-grounded agent and authority-aware benchmark for drug-information question answering. arXiv. 2026; arXiv:2606.01434. https://arxiv.org/abs/2606.01434
  114. ```
  115. ## Read More
  116. - Repository guide: `docs/repository-guide.md`
  117. - Maintainer guide: `maintainers/README.md`

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.