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 genomoncology/biomcp --skill biomcp-researchgit clone --depth 1 https://github.com/genomoncology/biomcpWrote 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/genomoncology/biomcp/biomcp-research)<a href="https://agentmods.dev/skills/genomoncology/biomcp/biomcp-research"><img src="https://agentmods.dev/badge/skills/genomoncology/biomcp/biomcp-research/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/genomoncology/biomcp/biomcp-research"><img src="https://agentmods.dev/badge/skills/genomoncology/biomcp/biomcp-research.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 | $0.00036 | $0.01549 |
| Opus 5 | $0.00018 | $0.00775 |
| Sonnet 5 | $0.00007 | $0.00310 |
| Haiku 4.5 | $0.00004 | $0.00155 |
Grade A, and why
biomcp-research 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 13d 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.
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 — 137 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Research with BioMCP
biomcp is on PATH; source is repos/biomcp. biomcp list is the
command reference, -j/--json for machine-readable output.
Two jobs run together here. Answer the question you were asked, and notice what the tool could not do while answering it. The second is not a distraction — a research session is the only time anyone finds out where the gaps are, and if they are not written down that day they are lost.
The loop
- Web search first, breadth. Find out what exists — the paper, the registry entry, the guideline. BioMCP is precise, not exploratory; it answers about things you can already name.
- BioMCP for the record.
get article <pmid>for the citable metadata and abstract,search article/article citations/article referencesto walk outward from a seed. - Go to the primary source. If the answer lives in a registry, a database, or a specification, fetch it directly. Quote it.
- Prefer practice over prose. When a document contradicts itself, what people actually did settles it better than what the document meant. Say which one you used.
Context discipline
get article <id> fulltext prints a path, not the text. That is
deliberate: a paper is tens of kilobytes and dumping it pins that
cost into the session for good. Do not reflexively read the whole
file back.
Read the abstract from get article <id> first — it answers more
questions than you expect. If you must open the cached file, grep it
for the term you need, or read a bounded range. Read it whole only
when you are genuinely going to use most of it.
Same for --json output. Pipe it through a filter; do not print
manifests to the transcript to look at one field.
Verify before you trust it
Cross-entity data is joined from several upstreams and the joins can
be wrong. A protein change and a cDNA change on the same line can
come from different transcripts. A .xlsx that arrives as
text/html is a download placeholder, not a spreadsheet.
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.
- 13d ago First seen · 137 lines · 36 tokens per session scan A 37297b04449b
biomcp-research is a skill published in the GitHub repository genomoncology/biomcp (630 stars, last pushed today), licensed MIT. It adds 36 tokens to every session and 1,549 once invoked, about $0.0002 per session on Opus 5. 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
biological-expert
Expert-level biology, biotechnology, genetics, bioinformatics, and computational biology. Use when the user mentions biology, biotechnology, genetics, bioinformatics, or genomics, or when the task involves Molecular Biology, Genomics & Bioinformatics, Systems Biology, or Data Analysis.
clinical-trials-search
Search ClinicalTrials.gov with natural language queries. Find clinical trials, enrollment, and outcomes using Valyu semantic search.
biopython
Comprehensive molecular biology toolkit. Use for sequence manipulation, file parsing (FASTA/GenBank/PDB), phylogenetics, and programmatic NCBI/PubMed access (Bio.Entrez). Best for batch processing, custom bioinformatics pipelines, BLAST automation. For quick lookups use gget; for multi-service integration use…
scanpy
Standard single-cell RNA-seq analysis pipeline. Use for QC, normalization, dimensionality reduction (PCA/UMAP/t-SNE), clustering, differential expression, and visualization. Best for exploratory scRNA-seq analysis with established workflows. For deep learning models use scvi-tools; for data format questions use…
structure-prediction
Protein structure prediction from sequence. ESMFold-based, single GPU, no MSA needed. Predicts 3D structures with pLDDT confidence scores for drug discovery targets.
cellxgene-census-query
Query CZ CELLxGENE Census (61M+ cells). Filter by cell type/tissue/disease, retrieve expression data, and integrate with scanpy/PyTorch for population-scale single-cell analysis. Use this skill when: (1) Querying single-cell expression data by cell type, tissue, or disease, (2) Exploring available single-cell datasets…