1. Set up the connection

Reuse an existing connection, or connect your agent below.

Where will you use Biosimulant?

Already connected? Create a session

2. Start a recorded investigation

Work in the assistant you already use. Studio keeps the plans, runs, results and evidence saved through Biosimulant together so you can review and continue them.

  1. Open Studio, select New and enter what you want your agent to investigate.
  2. Choose your app, create a session, then copy its prompt and send it in your connected assistant. It tells the agent which session to use.
  3. Follow the saved work in Timeline, Results and Evidence. Review proposed changes and runs before approving them.
  4. Return to the session and select Continue in your agent to give the same investigation to your assistant or another connected agent.

You can start with a research question, before you need a Lab or compute. Your assistant's full conversation and work through other tools are not copied automatically. Ask it to save supporting material and check where it was saved.

Follow the session quickstart

When your question needs a Lab

These recorded examples show model discovery, execution and compatibility checks.

Find a model and work out what it is

See the actual reply
Find the Bexkens 2024 S. mansoni glycolysis lab on the Hub and explain what it models.

Founddemi/bexkens2024-schistosoma-mansoni-glycolysis@1.0.0

SHA-256ea0fed3fd767794455b514a2d8a5d74200c595b38850b3e01b34367158ac4843

A kinetic model of glycolysis in S. mansoni, focused on how allosteric regulation of lactate dehydrogenase affects glycolytic behaviour.

Paper10.1016/j.ijpara.2024.03.005 · PMID 38492780

Qualification: schema_valid: true, but runtime_valid: false. Treat it as a package with a valid schema, not a domain-validated reproduction.

Copy it and run it

See the actual reply
Fork it into a workspace and run it with the default inputs.

Waiting for your approval

Once you allow it:

Runcompleted · duration 10.0 · step 1.0

Resultsresults/default_run_results.json

Of the selected observables, AMP moved most across the window. ATP, ADP, pyruvate, NAD and NADH were reported too.

Check two models fit together

See the actual reply
Can I feed this protein sequence into the Boltz affinity model?

protein_sequence declares protein.sequence/v1. So does the port you want to wire it to, so the connection holds.

ligand_smiles declares chemical.smiles/v1. Wiring a sequence into that one fails the check before anything runs.

3. Nothing runs until you say so.

Review the proposed Lab changes, run setup and publication before approving them. Reading saved work or starting a research session does not start compute.

  1. prepare

    The agent works out exactly what it wants to do. Nothing changes yet.

  2. You read it

    You see the exact plan, and which version of the model it will use.

  3. You approve

    Or you do not, and it stops there.

  4. apply

    Only now does anything run, save or publish.

Other things you might need

Keeping files off the cloud

For files that have to stay on your own machine. Work you do there is not saved as a version here, and gets no Passport.

Install the CLI

Skills

Two documents for writing down what a model has to do before you build it. Instructions, not tools: no extra access, no extra permissions. Most people never need them. A connected agent already reads the current versions from the server, so a download is only for keeping a copy of your own.

Download skills

Resume work across assistants

Keep the saved work around an investigation in one session. Each teammate can use their own agent in a shared team workspace. Review the record before continuing; another agent does not inherit the previous agent's execution approvals.

Read the session guide

The server address is https://api.biosimulant.com/mcp. For permissions, troubleshooting and the full tool list, read the agent guide (opens in a new tab)