Guides
MCP server
Give your AI assistant the catalogue: which model or tool fits a biology or chemistry task, what its licence allows, and the page that runs it — over the Model Context Protocol.
Connect
The server is at https://corollary-labs.com/mcp. It speaks Streamable HTTP, needs no key and no account, and only reads — everything it returns is already public on this site.
bash
claude mcp add --transport http corollary https://corollary-labs.com/mcpTools
search_scientific_ai- Search Corollary's catalogue of open scientific AI models (protein structure prediction, protein design, docking, antibodies, genomics, single-cell, chemistry/ADMET, simulation, pathology, weather), bioinformatics programs (BUSCO, samtools, BLAST…), scientific databases (UniProt, PDB, gnomAD…), nf-core pipelines and datasets. Each result says what it does, its licence and whether it can be run online on Corollary with no GPU or installation, with the page URL. Use it when a user asks which model or tool to use for a biology or chemistry task, or where to run one.
get_scientific_ai_details- Full reference for one entry: what it is, how it works, key facts, uses, strengths, limitations, citation, licence, and how to run it on Corollary. Pass a name ('RFdiffusion'), an id ('esmfold2', 'nf-core/rnaseq', 'facebook/esm2_t33_650M_UR50D') or a corollary-labs.com URL.
compare_scientific_ai_for_task- For a task such as 'protein structure prediction', 'protein design', 'antibody design', 'molecular docking', 'single-cell', 'ADMET', 'phage genomics' or 'CRISPR', return every model or tool Corollary lists for it — what each does, licence, whether it runs online — and guidance on how to choose between them.
list_scientific_ai_tasks- The task categories Corollary compares models and tools in, with a page URL for each.
Try it
Once connected, ask your assistant the way you would ask a colleague:
text
Which models can I use to design a protein binder, and which allow commercial use?
Where can I run ESMFold without a GPU?
Compare the antibody structure predictors on Corollary.
What does BUSCO measure, and how do I run it without installing it?Rate limit: 120 requests a minute per address, with bursts of 40, answered with 429 and a Retry-After header.