Practical overview
When it fits
Use Agents when the goal is to train or refine a language-agent pipeline based on task outcomes and recorded execution details, rather than only execute a fixed agent workflow.
Who it is for
- Developers building language-agent pipelines with prompts and tools.
- Researchers or engineering teams evaluating agent outcomes and iterating on pipeline components.
- Teams experimenting with multi-agent systems represented as pipeline nodes or multiple agents acting within a node.
Documented uses
- Run a language-agent pipeline on training examples and capture its execution trajectory.
- Assess pipeline results with a prompt-based loss.
- Use language-based reflections and gradients to update prompts, tools and pipeline connections.
- Experiment with optimizing multi-agent arrangements.
Requirements & limits
The repository is public, describes the framework as open source and lists the Apache-2.0 license. It documents installation from the master branch and local editable installation.
- The learning process depends on a designed agent pipeline and a prompt-based loss that evaluates outcomes.
- The framework records detailed trajectories, including prompts and tool usage, to inform updates.
- Updates can affect both symbolic components within nodes and the pipeline graph connecting them.
- The README documents installation routes, while its learning and evaluation overview points to separate project documentation for further use.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Install Agents from the master branch
Install the framework from its public Git repository using the documented pip command.
Before you start
- Python environment with pip
- Git access to https://github.com/aiwaves-cn/agents
Steps
Run the command below.
pip install git+https://github.com/aiwaves-cn/agents@masterContinue with the project documentation for learning and evaluation workflows.
Expected outcome
The repository's master branch is installed as a Python package.
Guide limitations
- This installation instruction does not document a training run or provide environment-specific compatibility details.
Install Agents for local development
Clone the master branch and install the checkout in editable mode.
Before you start
- Python environment with pip
- Git installed
Steps
Run the command below.
git clone -b master https://github.com/aiwaves-cn/agentsChange into the
agentsdirectory withcd agents.Run the command below.
pip install -e .Use the project documentation for the next stage, including learning and evaluation.
Expected outcome
The cloned repository is installed in editable mode for local development.
Guide limitations
- The README installation steps do not describe a training run or environment-specific compatibility details.
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 14 October 2026
- Original website title
- aiwaves-cn/agents · Official homepage title · retained in its source language