A general-purpose biomedical AI agent capable of executing complex research workflows using over 150 tools and databases.
Biomni is a "General-Purpose" biomedical agent. Unlike specialized skills (e.g., just for folding proteins), Biomni acts as a high-level orchestrator that can break down complex open-ended research questions into solvable sub-tasks using a vast library of tools.
When to Use This Skill
Complex Questions: "What are the potential drug targets for Alzheimer's related to mitochondrial dysfunction?"
Multi-Step Research: Tasks requiring literature search, data retrieval, and analysis in sequence.
Exploratory Analysis: When the exact path to the answer isn't known and requires "reasoning."
Core Capabilities
Tool Use: Access to 150+ tools, 105 software packages, and 59 databases.
Hypothesis Generation: Can formulate scientific hypotheses and plan experiments to test them.
Plan & Execute: Breaks down queries into a dependency graph of tasks.
Database QA: High accuracy (74.4%) in querying biomedical databases.
Workflow
Query Parsing: The agent analyzes the user's natural language request.
Tool Selection: It selects relevant tools (e.g., "Search GWAS Catalog", "Run GO Enrichment", "Fetch Uniprot Data").
Execution Loop:
Step 1: Get data.
Step 2: Analyze data.
Step 3: Refine plan based on intermediate results.
Synthesis: Combines all findings into a comprehensive answer.
Example Usage
User: "Identify genes associated with Type 2 Diabetes that are also expressed in the pancreas and have approved drugs."
Agent Action:
Tool: OpenTargets -> Get T2D associated genes.
Tool: GTEx -> Filter for high expression in Pancreas.
Tool: DrugBank -> Intersect with drug targets.
Result: Returns a list of genes (e.g., GLP1R, DPP4) and their drugs.
1---2name: biomni-research-agent3description: <!--4---5<!--6# COPYRIGHT NOTICE7# This file is part of the "Universal Biomedical Skills" project.8# Copyright (c) 2026 MD BABU MIA, PhD <md.babu.mia@mssm.edu>9# All Rights Reserved.10#11# This code is proprietary and confidential.12# Unauthorized copying of this file, via any medium is strictly prohibited.13#14# Provenance: Authenticated by MD BABU MIA1516-->1718---19name: biomni-research-agent20description: Bio-Research Generalist21license: MIT22metadata:23 author: Stanford (Snap Lab)24 source: "https://github.com/snap-stanford/Biomni"25 version: "1.0.0"26compatibility:27 - system: Python 3.9+28allowed-tools:29 - run_shell_command30 - web_fetch31 - python_repl32keywords:33 - biomni34 - automation35 - biomedical36 - reasoning37 - tools38measurable_outcome: Execute complex research tasks with >95% success rate and validated tool usage.39---4041# Biomni (General Biomedical Agent)4243A general-purpose biomedical AI agent capable of executing complex research workflows using over 150 tools and databases.4445Biomni is a "General-Purpose" biomedical agent. Unlike specialized skills (e.g., just for folding proteins), Biomni acts as a high-level orchestrator that can break down complex open-ended research questions into solvable sub-tasks using a vast library of tools.4647## When to Use This Skill4849* **Complex Questions**: "What are the potential drug targets for Alzheimer's related to mitochondrial dysfunction?"50* **Multi-Step Research**: Tasks requiring literature search, data retrieval, and analysis in sequence.51* **Exploratory Analysis**: When the exact path to the answer isn't known and requires "reasoning."5253## Core Capabilities54551. **Tool Use**: Access to 150+ tools, 105 software packages, and 59 databases.562. **Hypothesis Generation**: Can formulate scientific hypotheses and plan experiments to test them.573. **Plan & Execute**: Breaks down queries into a dependency graph of tasks.584. **Database QA**: High accuracy (74.4%) in querying biomedical databases.5960## Workflow61621. **Query Parsing**: The agent analyzes the user's natural language request.632. **Tool Selection**: It selects relevant tools (e.g., "Search GWAS Catalog", "Run GO Enrichment", "Fetch Uniprot Data").643. **Execution Loop**:65 * Step 1: Get data.66 * Step 2: Analyze data.67 * Step 3: Refine plan based on intermediate results.684. **Synthesis**: Combines all findings into a comprehensive answer.6970## Example Usage7172**User**: "Identify genes associated with Type 2 Diabetes that are also expressed in the pancreas and have approved drugs."7374**Agent Action**:751. Tool: `OpenTargets` -> Get T2D associated genes.762. Tool: `GTEx` -> Filter for high expression in Pancreas.773. Tool: `DrugBank` -> Intersect with drug targets.784. Result: Returns a list of genes (e.g., *GLP1R*, *DPP4*) and their drugs.7980<!-- AUTHOR_SIGNATURE: 9a7f3c2e-MD-BABU-MIA-2026-MSSM-SECURE -->
Run npx skillmds@latest add biotender-max/biomni-research-agent in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.
<!-- It is listed under AI & ML on SkillMD.
This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.
This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.
Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.
BioTender-max (@biotender-max) published this skill. Their other Agent Skills are listed on their SkillMD profile.
BioTender-max