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 hannesill/m4 --skill m4-apigit clone --depth 1 https://github.com/hannesill/m4Wrote 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/hannesill/m4/m4-api)<a href="https://agentmods.dev/skills/hannesill/m4/m4-api"><img src="https://agentmods.dev/badge/skills/hannesill/m4/m4-api/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/hannesill/m4/m4-api"><img src="https://agentmods.dev/badge/skills/hannesill/m4/m4-api.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00000 | $0.01658 |
| Opus 5.5 | $0.00000 | $0.00663 |
| Sonnet 5.5 | $0.00000 | $0.00332 |
| Haiku 4.5 | $0.00000 | $0.00166 |
Grade A, and why
m4-api scanned grade A with 0 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 yesterday.
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.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 115 lines — stays where its author put it; the contents beside it link to each section on GitHub.
M4 Python API
Every call names its dataset; M4 keeps no active dataset.
Every query
- Check which datasets are ready:
m4 list-datasets --json("prepared": true) orm4 status --all.list_datasets()returns every dataset M4 knows, downloaded or not. Adding data (m4 init,m4 download) needs the user's data access, so ask them first. - Find the tables and columns, then query with schema-qualified names:
from m4 import M4Client
db = M4Client(dataset="mimic-iv-demo")
db.schema()["tables"] # ["mimiciv_hosp.admissions", ...]
db.table_info("mimiciv_icu.icustays")["schema"] # columns (name, type); ["sample"] holds rows
df = db.query("""
SELECT first_careunit, COUNT(*) AS stays
FROM mimiciv_icu.icustays
GROUP BY first_careunit
ORDER BY stays DESC
""") # pandas DataFrame
execute_query(sql, dataset=...), get_schema(dataset=...), and
get_table_info(table, dataset=...) do the same without a client. From the shell:
m4 query --dataset mimic-iv-demo --sql "SELECT ...".
A query returns every row it selects: aggregate in SQL and keep row-level results in DataFrames, not in the conversation.
Datasets
| Dataset | Schemas | Access |
|---|---|---|
mimic-iv-demo |
mimiciv_hosp, mimiciv_icu |
open, 100 patients |
mimic-iv |
mimiciv_hosp, mimiciv_icu; mimiciv_derived after m4 init-derived |
credentialed |
mimic-iv-note |
mimiciv_note, for the notes functions only |
credentialed |
mimic-iv-ed-demo |
mimiciv_ed |
open, the MIMIC-IV demo's 100 patients |
mimic-iv-ed |
mimiciv_ed |
credentialed |
eicu-demo |
eicu_crd, as in eicu |
open, over 2,500 unit stays |
eicu |
eicu_crd (eicu_crd.patient, never bare patient) |
credentialed |
mimic-iii |
mimiciii_clinical |
credentialed |
mimic-cxr-jpg |
mimic_cxr: metadata, split, chexpert, negbio, test_labels, images |
credentialed, with a DUA of its own |
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.
- yesterday Changed · +32 lines eef4cc6794d8
- 2d ago Changed · +5 lines ca6eb4cb31e6
- 3d ago Changed · -91 lines · -38 tokens per session a0f81fe08454
- 1mo ago Changed · +1 lines 4adec078945c
- 1mo ago First seen · 168 lines · 38 tokens per session scan A a6ae7b2497b3
m4-api is a skill published in the GitHub repository hannesill/m4 (46 stars, last pushed yesterday), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,658 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
tooluniverse-single-cell
Single-cell RNA-seq analysis with scanpy/anndata — h5ad data loading, scRNA-seq quality control and QC gating (ngenesbycounts, totalcounts, mitochondrial percent / pctcountsmt, pctcountsribo, doublet detection with Scrublet/scDblFinder, ambient RNA / SoupX awareness, empty-droplet filtering, MAD-based thresholds)…
tooluniverse-protein-structure-prediction
Protein 3D structure prediction from sequence — ESMFold de novo prediction, AlphaFold database retrieval, experimental structures from RCSB, ProtVar variant impact assessment, ProtParam sequence properties. Use for structure prediction when no experimental structure exists, fold-confidence scoring, and…
tooluniverse-protein-sae-variant-interpretation
Interpret a missense variant via ESMC-6B Sparse Autoencoder (SAE) feature activations. For a given protein + variant, computes which interpretable SAE features (catalytic, ligand-binding, PTM, structural motif, domain, etc.) are lost or gained at the mutation site. Use when standard pathogenicity scores…
tooluniverse-enzyme-design
De novo enzyme design by catalytic-motif scaffolding — build a transition-state motif, generate scaffolds with AlphaProtein Novo, fold every mechanistic state, and triage with mechanism-inspired filters (catalytic geometry, multi-state H-bonds, multi-resequence backbone scoring). Use for designing enzymes for a named…
tooluniverse-proteomics-analysis
Mass-spec proteomics analysis — protein identification, quantification (LFQ, TMT, iTRAQ), differential expression (tumor vs normal, treatment vs control), PTM identification, and pathway enrichment on protein lists. Use when you have proteomics MS output, asking about protein abundance differences, or doing…
tooluniverse-spatial-omics-analysis
Spatial multi-omics interpretation pipeline. Transforms spatially variable genes (SVGs), domain annotations, and tissue context into biological insights via domain-by-domain characterization, cell-type composition, spatial gene expression patterns, RNA+protein+metabolite integration. Use for Visium, MERFISH, seqFISH…