Skip to main content

Agent Library Overview

How to navigate the new Agent Library

S
Written by Satya Singh
Updated this week

The new Agent Library centralizes guided onboarding and four specialized AI agents—helping teams design protocols, automate workflows, import data, and visualize results all in one place.

Onboarding Agent

The Onboarding Agent provides step-by-step guidance to help new users get started quickly. It answers questions about protocols, sample tracking, and lab workflows, drawing from Scispot’s community articles and documentation to ensure a smooth onboarding experience. Choose from Wet Lab Agent or Dry Lab Agent.

Wet Lab Agent

The Wet Lab Agent supports scientists with protocol-related inquiries, sample management, and day-to-day lab workflows. It leverages Scispot’s knowledge base to provide precise, lab-ready guidance.

Dry Lab Agent

The Dry Lab Agent focuses on workflow automation and API integrations. Upload files or reference items directly in your prompt to receive:

  • Ready-to-run code

  • Structured markdown outputs

  • API examples tailored to your context

Execute Agent

The Execute Agent allows you to generate new protocols and Labsheets seamlessly.

Generate Protocols

  • Type a prompt to create a protocol from scratch

  • Or upload an existing SOP in PDF or .docx format (using the + button)

  • Review the preview, then choose to save as a template or save as a protocol

  • Regenerate the protocol at any time by entering a new prompt

Generate Labsheets

You can also create Labsheets directly from a CSV or XLSX file.

  1. Upload your spreadsheet

  2. Click Next, then enter a Labsheet name and optional description

  3. Review the auto-generated columns based on your file

  4. Adjust column names and types as needed, and add additional or magic columns

This workflow mirrors the “Import File” option in the Labsheet dashboard.

Analytics Agent

The Analytics Agent transforms Labsheets into charts based on natural-language prompts, making reporting and collaboration simple. Just follow these steps:

  1. Select the Labsheet you want to analyze

  2. (Optional) Add filters by selecting Add Filters

  3. Enter a prompt describing the chart you want

    • Specify the chart type and the number of charts

    • Choose between Fast (quick responses) or Expert (high accuracy) models

  4. View the generated chart in the preview panel

    • Zoom in by dragging your mouse

    • Zoom out by double-clicking

Once you've generated your chart preview, here are some additional options you may choose:

  • Code: View the exact Python Plotly code used to generate your chart

  • Export: Download the chart as a PNG or export the Python code

  • Publish:

    • Add to Labsheet (appears in the Graphing Dashboard)

    • Add to Dashboard (appears in your homepage Dashboard)

You can modify the chart at any time by entering another prompt or adjusting the filters, then regenerating the results.

Did this answer your question?