Hybrid-Engine Data Analysis
code-yeongyu/oh-my-openagent
Analyzes CSV, Parquet and JSON data with DuckDB, Polars, numpy and matplotlib, preferring a persistent kernel over repeated one-shot processes.
Guide for using OptimusKG, the biomedical knowledge graph, through the optimuskg Python client.
Install Optimuskg
$ npx skills add mims-harvard/OptimusKG --skill optimuskg -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install mims-harvard/OptimusKG optimuskg --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/mims-harvard/OptimusKG.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/optimuskg .claude/skills/optimuskg && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "optimuskg" agent skill from https://github.com/mims-harvard/OptimusKG/tree/main/skills/optimuskg into .claude/skills/optimuskg/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "optimuskg", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/mims-harvard/OptimusKG/tree/main/skills/optimuskgType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add mims-harvard/OptimusKG --skill optimuskg -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install mims-harvard/OptimusKG optimuskg --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/mims-harvard/OptimusKG.git skills-src && mkdir -p .agents/skills && cp -r skills-src/skills/optimuskg .agents/skills/optimuskg && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "optimuskg" agent skill from https://github.com/mims-harvard/OptimusKG/tree/main/skills/optimuskg into .agents/skills/optimuskg/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "optimuskg", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add mims-harvard/OptimusKG --skill optimuskg -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install mims-harvard/OptimusKG optimuskg --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/mims-harvard/OptimusKG.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/skills/optimuskg .cursor/skills/optimuskg && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "optimuskg" agent skill from https://github.com/mims-harvard/OptimusKG/tree/main/skills/optimuskg into .cursor/skills/optimuskg/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "optimuskg", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/mims-harvard/OptimusKG.git --path skills/optimuskg--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add mims-harvard/OptimusKG --skill optimuskg -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install mims-harvard/OptimusKG optimuskg --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/mims-harvard/OptimusKG.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/skills/optimuskg .gemini/skills/optimuskg && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "optimuskg" agent skill from https://github.com/mims-harvard/OptimusKG/tree/main/skills/optimuskg into .gemini/skills/optimuskg/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "optimuskg", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install mims-harvard/OptimusKG optimuskgInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add mims-harvard/OptimusKG --skill optimuskg -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/mims-harvard/OptimusKG.git skills-src && mkdir -p .github/skills && cp -r skills-src/skills/optimuskg .github/skills/optimuskg && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "optimuskg" agent skill from https://github.com/mims-harvard/OptimusKG/tree/main/skills/optimuskg into .github/skills/optimuskg/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "optimuskg", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add mims-harvard/OptimusKG --skill optimuskg -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install mims-harvard/OptimusKG optimuskg --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/mims-harvard/OptimusKG.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/skills/optimuskg .opencode/skills/optimuskg && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "optimuskg" agent skill from https://github.com/mims-harvard/OptimusKG/tree/main/skills/optimuskg into .opencode/skills/optimuskg/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "optimuskg", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
Facts
- Skill name
optimuskg- GitHub stars
- 147
- Token cost
- ~1.9k tokens
- SKILL.md length
- 581 words
- Files
- 2
- Skills in repo
- 4
- Repo updated
- First seen
- Licence
- MIT
At a glance
Guide for using OptimusKG, the biomedical knowledge graph, through the optimuskg Python client.
- Tasks that involve DataFrames
- SKILL.md covers When to use, Installation, Quick start and Choosing the right loader, plus 6 more sections
- Calls uv and pip; reaches dataverse.harvard.edu
- Tasks that involve Knowledge graphs
What it does
Optimuskg is an agent skill from mims-harvard/OptimusKG. Guide for using OptimusKG, the biomedical knowledge graph, through the optimuskg Python client. Use this when loading, querying, filtering, or analyzing OptimusKG data — genes, drugs, diseases, phenotypes, anatomy, pathways, and their relationships — as Polars DataFrames or a NetworkX graph, or when downloading the published graph from Harvard Dataverse.
Its SKILL.md is about 1.9k tokens, which your agent loads only when the skill is triggered. The skill folder holds 2 other files (for example `reference/graph-schema.md`).
It sits in Data & Analytics, covering DataFrames and Knowledge graphs. It works with NetworkX, Polars and Python. The repository describes itself as: A modern multimodal knowledge graph with type-specific metadata across biomedical domains. The licence is MIT.
When your agent uses it
- Tasks that involve DataFrames
- Tasks that involve Knowledge graphs
Example prompts
- “/optimuskg”
Requirements
- Python 3
What it can do on your machine
Read from SKILL.md and the folder at commit 4fb3529. It shows what the files ask for, not the result of running them.
Tool permissions
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Runs code
Shell commands in SKILL.md call:
uvpip
From the folder's file list and the shell code blocks in SKILL.md.
Network
Hosts in commands or code, which the agent is likely to contact:
dataverse.harvard.edu
Also links to:
optimuskg.aidoi.org
From URLs in SKILL.md, links to its own repository left out.
Credentials
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Context cost
Optimuskg loads about 1.9k tokens when it runs. Until then it costs about 92 tokens; SKILL.md has 581 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
Safety
Auto-check passed
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.
SKILL.md
The full file from mims-harvard/OptimusKG at commit 4fb3529, republished under its MIT licence (© mims-harvard). 581 words, ~1,872 tokens.
Download SKILL.mdSave it as
.claude/skills/optimuskg/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.- name
- optimuskg
- description
- Guide for using OptimusKG, the biomedical knowledge graph, through the `optimuskg` Python client. Use this when loading, querying, filtering, or analyzing OptimusKG data — genes, drugs, diseases, phenotypes, anatomy, pathways, and their relationships — as Polars DataFrames or a NetworkX graph, or when downloading the published graph from Harvard Dataverse.
OptimusKG
OptimusKG is a modern multimodal biomedical knowledge graph (190,531 nodes
across 10 entity types, 21,813,816 edges across 27 relation types) integrating
65 resources grounded in 18 ontologies via the BioCypher framework and Biolink
Model. It is published on Harvard Dataverse
as Apache Parquet files and consumed through the optimuskg PyPI client.
This skill covers using the published graph via the client. It is not for
developing the data pipeline in the mims-harvard/optimuskg repo — that work
uses the repo's own /node-catalog-sync skill.
When to use
Use the optimuskg client when you need to:
- Download OptimusKG node/edge tables for analysis.
- Load the graph as Polars DataFrames or a NetworkX
MultiDiGraph. - Filter by entity type (Gene, Drug, Disease, …) or relation type.
- Build biomedical analyses, Graph-RAG, or ML over the graph.
Always prefer this client over hand-rolling Dataverse downloads — it resolves file IDs automatically and caches locally.
Installation
uv add optimuskg # in a uv project (preferred — see the `uv` skill)
pip install optimuskgThe client depends on polars (DataFrames) and networkx (graph view).
Quick start
import optimuskg
# Download a specific file; returns its local cached path
local_path = optimuskg.get_file("nodes/gene.parquet")
# Read a single Parquet file as a Polars DataFrame
drugs = optimuskg.load_parquet("nodes/drug.parquet")
# Load nodes + edges as Polars DataFrames (lcc=True -> largest connected component only)
nodes, edges = optimuskg.load_graph(lcc=True)
# Load as a NetworkX MultiDiGraph with properties merged onto node/edge attrs
G = optimuskg.load_networkx(lcc=True)Choosing the right loader
| Need | Function | Returns |
|---|---|---|
| Just the file on disk | get_file(path, *, force=False) | pathlib.Path |
| One table as a DataFrame | load_parquet(path, *, force=False, **read_parquet_kwargs) | pl.DataFrame |
| Whole graph as DataFrames | load_graph(*, lcc=False, force=False) | (nodes, edges) tuple of pl.DataFrame |
| Whole graph as NetworkX | load_networkx(*, lcc=False, force=False, parse_properties=True) | nx.MultiDiGraph |
Notes:
load_parquetforwards extra kwargs topl.read_parquet— e.g. push down column selection:optimuskg.load_parquet("nodes/drug.parquet", columns=["id"]).force=Truere-downloads even if the file is already cached.load_networkxalways builds aMultiDiGraphregardless of edge directionality; callG.to_undirected()if you need an undirected view.- Memory: loading the full graph into NetworkX needs several GB (190k
nodes, 21M edges) and emits a warning. Pass
lcc=Truefor a smaller, connected variant unless you specifically need every node.
File paths
Paths mirror the catalog layout under data/gold/kg/parquet/ in the source repo:
optimuskg.get_file("nodes.parquet") # full nodes table
optimuskg.get_file("edges.parquet") # full edges table
optimuskg.get_file("largest_connected_component_nodes.parquet") # LCC nodes
optimuskg.get_file("largest_connected_component_edges.parquet") # LCC edges
optimuskg.get_file("nodes/gene.parquet") # only Gene (GEN) nodes
optimuskg.get_file("edges/disease_gene.parquet") # only DIS-GEN edgesnodes/<type>.parquet files use the lowercase entity name (gene, drug, …);
edges/<a>_<b>.parquet files use the lowercase node pair (disease_gene,
drug_gene, …).
Show full SKILL.md (277 more words)Show less
Graph schema (at a glance)
Node table columns: id, label (the type code, e.g. GEN), properties.
Edge table columns: from, to, label (e.g. DIS-GEN), relation
(e.g. ASSOCIATED_WITH), undirected, properties.
In the unified nodes.parquet / edges.parquet tables, properties is a JSON
string. In the stratified per-type files (nodes/<type>.parquet,
edges/<label>.parquet) it is expanded into native typed columns as a Polars
Struct.
The 10 node type codes:
| Label | Type | Label | Type | |
|---|---|---|---|---|
GEN | Gene | ANA | Anatomy | |
DIS | Disease | MFN | Molecular Function | |
BPO | Biological Process | CCO | Cellular Component | |
PHE | Phenotype | PWY | Pathway | |
DRG | Drug | EXP | Exposure |
For the full edge-label/relation taxonomy (all 27 edge types and their relation
strings) and per-type property fields, see
reference/graph-schema.md.
Common patterns
Filter Polars DataFrames by type/relation:
import polars as pl
nodes, edges = optimuskg.load_graph(lcc=True)
genes = nodes.filter(pl.col("label") == "GEN")
dis_gen = edges.filter(pl.col("relation") == "ASSOCIATED_WITH")Filter a NetworkX graph (properties are merged onto attrs):
G = optimuskg.load_networkx(lcc=True)
# Nodes by type code
genes = [n for n, a in G.nodes(data=True) if a["label"] == "GEN"]
# Edges by relation
expression = [
(u, v) for u, v, a in G.edges(data=True)
if a["relation"] == "EXPRESSION_PRESENT"
]Pass parse_properties=False to load_networkx to keep properties as a raw
JSON string instead of merging parsed keys into the attribute dicts.
Configuration
The client targets doi:10.7910/DVN/IYNGEV on https://dataverse.harvard.edu
by default, and caches downloads in platformdirs.user_cache_dir("optimuskg")
(~/.cache/optimuskg on Linux, ~/Library/Caches/optimuskg on macOS). Cache
keys include the dataset version, so a new release invalidates it automatically.
Override from code or via environment variables:
optimuskg.set_cache_dir("/data/optimuskg-cache")
optimuskg.set_doi("doi:10.7910/DVN/EXAMPLE") # target a different release
optimuskg.set_server("https://dataverse.example.org") # non-Harvard installation
# Read current settings
optimuskg.get_cache_dir(); optimuskg.get_doi(); optimuskg.get_server()export OPTIMUSKG_CACHE_DIR=/data/optimuskg-cache
export OPTIMUSKG_DOI=doi:10.7910/DVN/EXAMPLE
export OPTIMUSKG_SERVER=https://dataverse.example.orgCiting and license
- Cite OptimusKG when you use it in research — see
https://optimuskg.ai/docs/citation and the dataset DOI
10.7910/DVN/IYNGEV. - License: the OptimusKG codebase is MIT. The integrated source datasets keep their own licenses and terms of use, which may restrict redistribution or commercial use of a given graph subset — review each source's terms. See https://optimuskg.ai/docs/license.
Documentation
For full details (function signatures, per-type schemas, edge relations), read the official docs — they expose machine-readable text files:
- https://optimuskg.ai/llms.txt — index of all doc pages
- https://optimuskg.ai/llms-full.txt — full documentation in one file
- https://optimuskg.ai/docs/optimuskg-client/reference — API reference
- https://github.com/mims-harvard/optimuskg — source repository
© mims-harvard, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
Files
SKILL.md and 1 other file in skills/optimuskg of mims-harvard/OptimusKG.
- SKILL.md
- reference/graph-schema.md
Open the folder on GitHubat commit 4fb3529
Compare with similar skills
Optimuskg next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Optimuskg this skillmims-harvard/OptimusKG | 147 | — | ~1.9k | Automated safety check: Pass | MIT | |
| Hybrid-Engine Data Analysiscode-yeongyu/oh-my-openagent | 70k | — | ~1.4k | Automated safety check: Pass | Custom licence | |
| Narwhalsanam-org/metaxy | 124 | — | ~3.3k | Automated safety check: Pass | Apache-2.0 | |
| Optimize For GPUK-Dense-AI/scientific-agent-skills | 48k | 1 repos | ~3.4k | Automated safety check: Pass | MIT | |
| PolarsK-Dense-AI/scientific-agent-skills | 48k | 1 repos | ~3.3k | Automated safety check: Pass | MIT | |
| Ingesting Dataancoleman/ai-design-components | 525 | — | ~1.9k | Automated safety check: Pass | MIT |
Similar skills
Narwhals
anam-org/metaxy
Effectively use Narwhals to write dataframe-agnostic code that works seamlessly across multiple Python dataframe libraries.
Optimize For GPU
K-Dense-AI/scientific-agent-skills
GPU-accelerates scientific Python on NVIDIA hardware and verifies that the result is correct and faster.
Polars
K-Dense-AI/scientific-agent-skills
High-performance DataFrame library for Python ETL, analytics, and pandas migration.
Ingesting Data
ancoleman/ai-design-components
Data ingestion patterns for loading data from cloud storage, APIs, files, and streaming sources into databases.
Transforming Data
ancoleman/ai-design-components
Transform raw data into analytical assets using ETL/ELT patterns, SQL (dbt), Python (pandas/polars/PySpark), and orchestration (Airflow).
More from mims-harvard/OptimusKG
Scientific Visualization
mims-harvard/OptimusKG
Create publication figures with matplotlib/seaborn/plotly. An agent skill from mims-harvard/OptimusKG.
Impeccable
mims-harvard/OptimusKG
Create distinctive, production-grade frontend interfaces with high design quality.
Node Catalog Sync
mims-harvard/OptimusKG
Enforce synchronization between Kedro node files and catalog YAML files in the OptimusKG project.
Related
Categories
Questions about Optimuskg
What does Optimuskg do?
Guide for using OptimusKG, the biomedical knowledge graph, through the optimuskg Python client. Optimuskg is an agent skill from mims-harvard/OptimusKG. Guide for using OptimusKG, the biomedical knowledge graph, through the optimuskg Python client.
When should I use Optimuskg?
Optimuskg fits situations like: tasks that involve DataFrames; tasks that involve Knowledge graphs.
How do I install Optimuskg in Claude Code?
Run `npx skills add mims-harvard/OptimusKG --skill optimuskg -a claude-code`. Or copy the skill folder (skills/optimuskg in mims-harvard/OptimusKG) into .claude/skills/optimuskg in your project. Claude Code loads it when a task matches its description.
How do I install Optimuskg in Codex?
Run `npx skills add mims-harvard/OptimusKG --skill optimuskg -a codex`. Or copy the skill folder (skills/optimuskg in mims-harvard/OptimusKG) into .agents/skills/optimuskg in your project. Codex loads it when a task matches its description.
Can I use Optimuskg in Cursor, Gemini CLI or GitHub Copilot?
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add mims-harvard/OptimusKG --skill optimuskg -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/optimuskg, .gemini/skills/optimuskg, .github/skills/optimuskg and .opencode/skills/optimuskg in your project.
What does Optimuskg need to run?
Going by SKILL.md and its folder, Optimuskg needs the command-line tools its instructions call (uv and pip). Our summary lists: Python 3.
Does Optimuskg access the network?
SKILL.md names 3 domains. In commands or code: dataverse.harvard.edu; the agent is likely to contact it when it follows the instructions. As links in the text: optimuskg.ai and doi.org. This is read from the text; nothing was executed.
Is Optimuskg safe to install?
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.
What licence does Optimuskg use?
Optimuskg is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
How many tokens does Optimuskg use?
About 1.9k tokens (SKILL.md is roughly 7.5k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.
What are the alternatives to Optimuskg?
Skills that share tags, products or a category with Optimuskg: Hybrid-Engine Data Analysis (code-yeongyu/oh-my-openagent, 70k stars), Narwhals (anam-org/metaxy, 124 stars), Optimize For GPU (K-Dense-AI/scientific-agent-skills, 48k stars) and Polars (K-Dense-AI/scientific-agent-skills, 48k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Who maintains Optimuskg?
mims-harvard (a GitHub organization) maintains it in mims-harvard/OptimusKG, which has 147 GitHub stars. The repository holds 4 skills in this directory. The repository was last updated on September 21, 2026.
Source: mims-harvard/OptimusKG on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.