Skip to content
TaskWeaver logo

TaskWeaver

Verified

Code-first agent framework for stateful data analytics execution.

Autonomous AgentsAgent Frameworks 6.2kOpen source
View on GitHub
Updated 2026-06-15
TaskWeaver GitHub repository

What is TaskWeaver?

TaskWeaver is an open-source code-first agent framework designed to plan and carry out data analytics tasks by generating and running code while coordinating available plugins.

It works by interpreting requests as executable code, maintaining full conversation and execution histories, and operating in a stateful way that supports iterative data processing.

The framework targets users who need reliable coordination of analytics functions on structured data and prefer a code-centric workflow over pure text-based agents.

Capabilities

plan data analytics tasks
execute data analytics tasks
code-first agent orchestration
seamless workflow integration

What you can build with TaskWeaver

Tabular Data Processing

Run analytics on high-dimensional tables while keeping intermediate results in memory across steps.

Plugin-Coordinated Workflows

Combine multiple function plugins into a single plan that executes data tasks without losing state.

Local Model Deployment

Operate the agent with smaller locally hosted language models for analytics pipelines.

Install TaskWeaver

Quick start
# [optional to create conda environment]
# conda create -n taskweaver python=3.10
# conda activate taskweaver

# clone the repository
git clone https://github.com/microsoft/TaskWeaver.git
cd TaskWeaver
# install the requirements
pip install -r requirements.txt
  1. 1Obtain the all-in-one Docker image for a complete setup.
  2. 2Switch to container mode for code execution security.
  3. 3Configure shared memory and experience features if needed.
  4. 4Start the agent and submit a data analytics request.
  5. 5Review both chat and code execution history for results.

TaskWeaver: pros & cons

Pros

  • +Maintains code execution history and in-memory data
  • +Stateful plugin coordination for iterative tasks
  • +Strong support for complex tabular structures
  • +Works with local language models and Docker

Cons

  • Focused mainly on data analytics use cases
  • Requires comfort with code generation and execution
  • Additional setup needed for container or shared memory features
Did you find this helpful?

Frequently asked questions

Yes, it preserves both chat history and code execution history including in-memory data.

User reviews

Verified reviews from the community shape this listing's rating.

Loading reviews…

Sign in to review

Promote TaskWeaver

Add this badge to your website, or share the tool.

DFeatured on DhanasviTaskWeaver 0