Practical overview
When it fits
Use Astromesh when building a Python-based agent service and you want a runtime to handle agent configuration, model routing and workflow orchestration, with API exposure and additional capabilities such as retrieval and memory.
Who it is for
- Developers building AI copilots, such as developer assistants, support agents or internal knowledge assistants.
- Teams implementing document-processing, business-automation or API-orchestration workflows.
- Developers creating multi-agent services or exposing agents as programmable APIs.
Documented uses
- Define an agent's model, prompt, identity and orchestration pattern in YAML.
- Route agent roles to models and coordinate work using patterns such as ReAct, Pipeline or Supervisor.
- Add retrieval, memory or MCP tool support to an agent runtime.
- Run agents behind REST or WebSocket APIs.
Requirements & limits
The repository is public and identifies Astromesh as open source under the Apache-2.0 license. The documented quick start requires Python 3.12+ and uv.
- The documented implementation is Python-based and requires Python 3.12 or later.
- The quick start uses uv for dependency management.
- Agent configuration is declarative YAML, and the example specifies an Ollama model.
- The README describes REST and WebSocket API exposure and seven orchestration patterns.
- A Docker Compose development stack is documented with inference, database, cache and observability components.
Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.
Set up Astromesh and run a first local agent
Install the runtime, start its API, define the documented example agent and send it a request.
Before you start
- Python 3.12 or later
- uv package manager
- Ollama model provider and the example llama3.1:8b model
Steps
Install uv with pip, then clone the Astromesh repository and change into its directory.
Install dependencies with uv sync.
Start the runtime using uv run uvicorn astromesh.api.main:app --reload.
Create config/agents/my-agent.agent.yaml using the README example, which specifies an Ollama model and ReAct pattern.
POST a JSON query to the documented /v1/agents/my-agent/run endpoint on localhost:8000.
Expected outcome
A locally running agent API with the example agent configuration and a documented request path.
Guide limitations
- The guide's request example assumes the runtime is available at localhost:8000.
Compare options for Agent orchestration
These projects document a common task and share a project role. This is a source-based comparison, not an execution benchmark or a guarantee of interchangeability.
| Project | Documented purpose | Platform | Access and requirements |
|---|---|---|---|
| Astromesh | Define an agent's model, prompt, identity and orchestration pattern in YAML. | Self-hosted, Server | The repository is public and identifies Astromesh as open source under the Apache-2.0 license. The documented quick start requires Python 3.12+ and uv. |
| Agenite | Define typed tools with JSON Schema validation and error handling. | CLI | |
| Veldra | Create an agent with a policy, tools, skills, knowledge base and reasoning method. | Self-hosted | The documented deployment is self-hosted and uses a sign-in gate after a first-run administrator account is created. |
Sources & review
Source record details
- Source availability
- Active
- Sources checked
- 13 October 2026
- Original website title
- monaccode/astromesh · Official homepage title · retained in its source language