Co-Scientist
Co-Scientist is Seqera's AI assistant for bioinformatics. It builds, runs, and debugs Nextflow pipelines, manages your data, and works with your Seqera Platform resources.
You can use Co-Scientist in two places:
- In Seqera Platform: Open the Co-Scientist panel from any workspace page. Because Co-Scientist reads the page you are on, you can ask about the run, pipeline, or dataset in front of you without describing it first. See Co-Scientist in Seqera Platform.
- In the Seqera CLI: Run
seqera aiin your terminal to work in your local checkout with access to your Platform workspace. See Installation.
Both use the same assistant and the same Seqera Platform account. Skills, modes, command approval, and the Nextflow language server protocol (LSP) are available only in the CLI.
Get started
In Seqera Platform
- Sign in to Seqera Platform and open a workspace.
- Select Co-Scientist in the navigation to open the panel.
- Ask a question about the page you are on, such as why a run failed.
See Co-Scientist in Seqera Platform for page context, conversation history, and workspace scope.
In the Seqera CLI
-
Install the Seqera CLI:
npm install -g seqera -
Log in to Seqera:
seqera login -
Start your first session:
seqera ai
See Installation for prerequisites, updates, and development builds. Then see Quickstart to walk through your first session.
What you can do
Co-Scientist works across the pipeline lifecycle, from writing Nextflow code to running it on Seqera Platform.
Develop pipelines
Generate Nextflow configurations and pipeline schemas, convert scripts from other languages (WDL, R) to Nextflow, and discover over 1,000 nf-core modules with ready-to-run commands. Build reproducible Wave containers from conda or pip packages without writing a Dockerfile. The Nextflow language server detects errors as you edit and lets Co-Scientist navigate Nextflow, Python, and R files.
Run and debug on Platform
Launch, monitor, and debug pipeline runs with real-time status, logs, and run metrics. Browse cloud storage through data-links, manage datasets, generate upload and download URLs, and access reference genomes. Co-Scientist works with the compute environments, datasets, and workspaces your account can already access.
Sessions, skills, and modes
Ask in plain English in Seqera Platform or the CLI. In the CLI, use reusable skills exposed as slash commands in the / palette. Switch between build, plan, and goal modes to match execution, analysis, or long-running tasks. Resume earlier sessions with seqera ai -c, and organize workspace resources into projects using Platform labels.
Learn more
- Co-Scientist in Seqera Platform: Use the Co-Scientist panel in Seqera Platform
- Installation: Install, update, and configure the CLI
- Quickstart: Run your first Co-Scientist session in the CLI
- Authentication: Log in, log out, and manage sessions
- Use cases: Seqera CLI use cases
- Using Co-Scientist: Configure modes, sessions, skills, command approval, and more
- Coding Agents: Install Co-Scientist as a skill in your coding agent
- Skills: Built-in skills, slash commands, and session limits
- Troubleshooting: Troubleshoot common errors