xianyu-sheng/Xenon — 开源终端 AI 编程 Agent：以在线 Evidence Runtime 验证链（LLM 输出是 Claim、工具结果才是 Evidence）为核心，SWE-bench_Lite 官方评测 36.7…

Description and use

Here's a brief summary and explanation of the open-source project Xenon:

**Summary:** Xenon is an open-source terminal AI programming agent that enables the creation of repeatable agent workflows by integrating online evidence runtime validation, supporting seven reasoning paradigms, and multiple models. It helps turn isolated prompts and tools into standardized workflows, making it suitable for coordinating multi-agent workflows, adding tool-use pipelines, and standardizing agent memory.

**Value Proposition:** The value of Xenon lies in its ability to turn isolated prompts and tools into repeatable agent workflows, making it easier to integrate and automate tasks. This enables developers to create more efficient and scalable workflows, reducing the time and effort required to complete complex tasks.

**Practical Adoption Path:** To adopt Xenon, developers can start by exploring its features and capabilities, such as its support for multiple reasoning paradigms and models. They can then integrate Xenon into their existing workflows by using its API, SDK, or CLI. The project's documentation and community support can provide additional resources and guidance for developers looking to adopt Xenon.

**Production Readiness:** Xenon has a high production readiness score, indicating that it is suitable for serious piloting and potential production deployment. The project has recent activity, adoption, and ecosystem signals, which suggests that

Problem

xianyu-sheng/Xenon helps turn isolated prompts and tools into repeatable agent workflows.

Outcome

High for an OSS candidate: recent activity, adoption, and ecosystem signals are strong enough for a serious pilot.

Workflow

• coordinate multi-agent workflows
• add tool-use pipelines
• standardize agent memory

Evidence

开源终端 AI 编程 Agent：以在线 Evidence Runtime 验证链（LLM 输出是 Claim、工具结果才是 Evidence）为核心，SWE-bench_Lite 官方评测 36.7% 实例级通过率（可复现报告）。支持 7 种推理范式、MCP 工具、多模型。Python。

Source