Drug Discovery Agent Skills plugin

37 skills for computational drug discovery: from target selection and docking to ADMET and the clinical record.

  • 55 ★ on the repository
  • Last commit
  • MIT
Plugin

Extended access

Why this level

  • Skill scripts run on your machine and call external APIs with your keys.
  • Calculation results are for research and need expert verification.
All reasons and checks

k-dense-ai/drug-discovery-agent-skills

Install

Manual install

npx skills add K-Dense-AI/drug-discovery-agent-skills

Detects installed agent hosts and installs there. Alternative: `gh skill install K-Dense-AI/drug-discovery-agent-skills`.

Access and checks

How we review

Extended accessThe tool runs code or makes network calls and only reads external data.

  • Skill scripts run on your machine and call external APIs with your keys.
  • Calculation results are for research and need expert verification.

This is third-party code. Review the repository files before installing.

What it does

The set of 37 skills teaches a coding agent the tools computational chemists and biologists use: how to install them, which API to call, what parameters mean and where each one breaks. It covers target validation and human genetics, bioactivity and purchasable compounds, generative design and retrosynthesis, docking, free energy and dynamics, ADMET and dose projection, proteins, antibodies, degraders and oligonucleotides, and the clinical and regulatory record. Skills chain together: one request often pulls in two or three, for example `open-targets`, `depmap` and `chembl`. A table states each one's needs: runs locally, calls a public API with no key, requires a key or a GPU. The repository follows the Agent Skills and Agent Plugins 1.0.0 standards.

Who it is for. Computational chemists, bioinformaticians and medicinal scientists who hand an agent routine data retrieval and preliminary calculations.

Good fit when

  • You need to find which targets in asthma have genetic evidence and are small-molecule tractable.
  • A clean EGFR IC50 dataset is built from ChEMBL and the amount of discarded data is counted.
  • For an EGFR structure with an inhibitor, you check that the ATP site is fully resolved.
  • Before synthesis you check whether a compound series is already claimed in patents.

Not a fit when

  • You need clinical decisions for a patient, since the set is meant for research work.
  • You do not want to give an agent network and API key access.
  • No GPU is available, but docking and folding of large complexes are needed.

Example request

Which targets in asthma have genetic evidence, are small-molecule tractable and are not essential in all cells? Show the sources.

Limitations

The skills give instructions and the scripts run on your machine with access to files, environment variables and the network, so study the `SKILL.md` files of the tools you need and `docs/security-report.md` before installing. Some skills use public APIs with no key, others need your own keys, for example Adaptyv, ESM Forge, Rowan and Tamarind. Some tools, such as Boltz-2, practically need a GPU, and some cap the Python version. Have an expert verify calculation results, as this is not medical advice.

How to disable. Remove the installed skills folder with your environment's tooling, for example delete `~/.agents/skills/drug-discovery-agent-skills`. Remove API keys from environment variables.

FAQ

Which install options exist?

`npx skills add K-Dense-AI/drug-discovery-agent-skills`, `gh skill install` with a version pin, a plugin for Cursor or Codex, and a plain `git clone` into the skills folder.

What does the Needs column mean?

It shows what a skill requires: a local run, a public API with no key, your own key or a GPU, plus Python version limits.

How do I pin a version?

With `gh skill install K-Dense-AI/drug-discovery-agent-skills --pin v1.3.0` or by cloning the `v1.3.0` tag.

Are API keys needed?

Not for some skills: public databases such as ChEMBL and Open Targets need no key. Others, such as Adaptyv and Rowan, need your own keys.

Does it work in other agents?

Yes, the authors name Claude Code, Cursor, Codex and Google Antigravity. The repository is packaged as Agent Plugins 1.0.0.

Discussion

No questions or reviews yet.

Plugins

Around 200 skills for Claude Code on bioinformatics, chemistry and biostatistics, from RNA-seq analysis to molecular docking

SkillExtended access388Repository stars
Editors’ pick

166 skills for scientific work: bioinformatics, cheminformatics, clinical data, geospatial analysis and 100+ databases

SkillExtended access57KRepository stars

Skills for computational chemistry: ASE, DFT codes, LAMMPS, DeePMD, RDKit, Open Babel and more, built for OpenClaw

SkillExtended access156Repository stars

17 skills for computational biology: RNA-seq, single-cell, protein structure, docking, literature and figures

SkillExtended access75Repository stars

A library of agent skills for medical research: literature search and appraisal, study design, data analysis and academic writing

SkillExtended access2KRepository stars

README in short

The README lists 37 skills by group with their requirements, three install routes, example questions, a security warning and version 1.3.0.

README badge

Report a problem