agentskillexchange Fork and merge Pydantic AI agent runs with Pydantic Deep Agents
Run a self-hosted terminal agent or custom Pydantic AI harness that can branch a coding or research run, test isolated alternatives, and merge the winning branch.
Prerequisites
Python 3.10+, pydantic-deep CLI or Python package, approved model provider key, optional sandbox support for isolated execution
Installation
Install the package in the Python environment that will run the local agent harness:
pip install pydantic-deep
Use the upstream documentation for current CLI and framework setup details:
A minimal ASE-safe workflow is:
- Install the package and CLI in an isolated Python environment.
- Configure only the approved model provider credentials needed by your harness.
- Run a branch or fork workflow from a self-hosted terminal agent or custom Pydantic AI agent.
- Compare branch outputs with tests, traces, or review notes.
- Merge or select the winning branch only after a human reviews the evidence.
Treat this as a self-hosted custom-agent workflow for branching and evaluating Pydantic AI runs. It is not a generic Claude Code replacement, and it should not be wired to broad shell or repository permissions until the branch/test/merge loop is constrained and reviewed.