ToolUniverse is a collection of tools, interfaces, and supporting components for building AI systems that perform scientific work. It is for developers creating AI scientist agents that use APIs, databases, machine-learning tools, and domain-specific utilities. The catalogue includes skills, commands, an MCP server, an agent, and a hook for working with the ecosystem.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add mims-harvard/ToolUniverse --skill tooluniverse-data-wranglinggit clone --depth 1 https://github.com/mims-harvard/ToolUniverseWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/mims-harvard/tooluniverse/tooluniverse-data-wrangling)<a href="https://agentmods.dev/skills/mims-harvard/tooluniverse/tooluniverse-data-wrangling"><img src="https://agentmods.dev/badge/skills/mims-harvard/tooluniverse/tooluniverse-data-wrangling/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/mims-harvard/tooluniverse/tooluniverse-data-wrangling"><img src="https://agentmods.dev/badge/skills/mims-harvard/tooluniverse/tooluniverse-data-wrangling.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 3 findings, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Data Exfiltration · line 195 Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
- medium Data Exfiltration · line 195 Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
- medium Data Exfiltration · line 195 Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00087 | $0.04626 |
| Opus 5 | $0.00044 | $0.02313 |
| Sonnet 5 | $0.00017 | $0.00925 |
| Haiku 4.5 | $0.00009 | $0.00463 |
Grade B, and why
tooluniverse-data-wrangling scanned grade B with 2 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Sends data to an external URLmediumData exfiltration
A POST to an outside endpoint may be telemetry or may be exfiltration; either way the mod talks to somewhere, and you should know where.
cases = requests.post("https://api.gdc.cancer.gov/cases", json={ Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
ids = requests.get(f"{base}/esearch.fcgi?db=gene&term=BRCA1+AND+human&retmax=500&retmode=json").json() How it starts
The opening of the file, as written. The whole thing — 399 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Data Wrangling: Universal Access Patterns
Reference for downloading and parsing scientific data from any source. Write and run Python code via Bash for every step.
When to Use
- ToolUniverse tool returned metadata/search results but you need raw or bulk data
- Data is in a format tools don't parse (VCF, h5ad, BAM, SDF, GCT)
- You need a multi-step API workflow (search -> filter -> download -> parse)
- The data source has no ToolUniverse tool at all
- You need thousands of records, not the 10-100 a tool returns
Decision: Tool vs Code
| Situation | Use |
|---|---|
| Single record lookup, simple search, <100 results | ToolUniverse tool (execute_tool) |
| Bulk download, custom filtering, format conversion | Write Python code |
| Tool exists but returns truncated results | Write code using the same API the tool wraps |
| No tool exists for this source | Write code directly |
Section A: Format Cookbook
Tabular
import pandas as pd, io
df = pd.read_csv("data.csv") # CSV
df = pd.read_csv("data.tsv", sep="\t") # TSV
df = pd.read_sas(io.BytesIO(content), format="xport") # SAS Transport (XPT) — NHANES, CDC
df = pd.read_sas("data.sas7bdat", format="sas7bdat") # SAS native
df = pd.read_stata("data.dta") # Stata — ICPSR, HRS
df = pd.read_parquet("data.parquet") # Parquet — MIMIC-IV
df = pd.read_excel("data.xlsx") # Excel
df = pd.read_spss("data.sav") # SPSS
df = pd.read_fwf("data.dat") # Fixed-width — legacy surveys
Genomics
from Bio import SeqIO
records = list(SeqIO.parse("seqs.fasta", "fasta")) # FASTA
records = list(SeqIO.parse("reads.fastq", "fastq")) # FASTQ
# VCF (no cyvcf2 needed)
vcf_lines = [l for l in open("vars.vcf") if not l.startswith("##")]
df = pd.read_csv(io.StringIO("".join(vcf_lines)), sep="\t")
df = pd.read_csv("genes.gff3", sep="\t", comment="#", # GFF/GTF
names=["seqid","source","type","start","end","score","strand","phase","attrs"])
df = pd.read_csv("regions.bed", sep="\t", header=None, # BED
names=["chrom","start","end","name","score","strand"])
import pysam # BAM (requires pysam)
bam = pysam.AlignmentFile("aligned.bam", "rb")
for read in bam.fetch("chr1", 1000, 2000): print(read.query_name)
What ships with it
1 file beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 10d ago First seen · 399 lines · 87 tokens per session scan B 09181d900dff
tooluniverse-data-wrangling is a skill published in the GitHub repository mims-harvard/ToolUniverse (1,678 stars, last pushed today), licensed Apache-2.0. It adds 87 tokens to every session and 4,626 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it B with 2 findings (sends data to an external url, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
dfam-check
Measure mesh files against Design for Additive Manufacturing (DfAM) rules and report printability findings per process (FDM, SLS, SLA/DLP, metal PBF, MJF). Use when the user asks whether a part is printable, wants overhang/wall-thickness/support analysis of an .stl, .obj, .ply, or .3mf mesh, wants a build-orientation…
nanoresearch-writing
Draft a LaTeX research paper from all previous stage outputs.
construct-toy-examples
Generate and analyze simpler examples that satisfy both the assumptions and the conclusion of a theorem statement or subgoal. Use when you are stuck in reasoning and need simpler examples to regain traction, or when you want to see where the assumptions take effect and gain intuition.
obtain-immediate-conclusions
Derive immediate mathematical consequences from a theorem statement or subgoal. Use when starting a new problem, branch, or subgoal, or when cheap progress or a cleaner reformulation is needed before deeper proof search.
astro-dso-doc
Generates a complete, polished HTML documentation page, a processing checklist, an AstroBin post JSON, a PixInsight process icon set (XPSM), AND a ready-to-paste PixInsight project Description field for a deep-sky object (DSO) astrophotography project. Use this skill whenever the user mentions astrophotography, a DSO…
intermediate-outputs
Use this skill when working with circuit discovery in language models, mechanistic interpretability, activation patching, attribution patching, or Layer-wise Relevance Propagation (LRP) for neural network analysis.