feat: support custom headers in LLM runner (EVOLVER_LLM_HEADERS)

[chliuqi]

问题

内置的 llm runner（llmRunner.js）在 complete() 函数中硬编码了 HTTP 请求头：

• "" + + "" + + "javascript
  headers: { 'content-type': 'application/json', authorization: Bearer \ },
• "" + + "" + + "" +

某些 OpenAI 兼容的 API 提供商需要额外的请求头来进行路由/认证。例如，OpenCode Go 需要 + "" + x-opencode-session + "" + 头来进行请求路由：

• "" + + "" + + "json
  {
  "error": {
  "type": "MissingSessionID",
  "message": "Request is missing x-opencode-session and cannot be routed efficiently."
  }
  }
• "" + + "" + + "" +

没有这个头，API 会返回 400 Bad Request。

建议方案

通过环境变量 + "" + EVOLVER_LLM_HEADERS + "" + 支持自定义请求头：

简单格式（单个头）：

• "" + + "" + + "�ash
  export EVOLVER_LLM_HEADERS='x-opencode-session:evolver-autoexec-12345'
• "" + + "" + + "" +

JSON 格式（多个头）：

• "" + + "" + + "�ash
  export EVOLVER_LLM_HEADERS='{"x-opencode-session":"evolver-12345","user-agent":"my-agent/1.0"}'
• "" + + "" + + "" +

实现建议

在 + "" +
eadLlmRunnerConfig() + "" + 中解析 + "" + EVOLVER_LLM_HEADERS + "" + 并合并到请求头：

• "" + + "" + + "javascript
  // readLlmRunnerConfig
  const rawHeaders = env['EVOLVER_LLM_HEADERS'] ?? '';
  const extraHeaders = rawHeaders.startsWith('{')
  ? JSON.parse(rawHeaders)
  : Object.fromEntries(rawHeaders.split(',').map(h => h.split(':').map(s => s.trim())));

return { apiKey, baseUrl, model, maxTurns, maxFileBytes, extraHeaders };

// complete()
const response = await fetchFn(\/chat/completions, {
method: 'POST',
headers: {
'content-type': 'application/json',
authorization: \Bearer \,
...config.extraHeaders,
},
body: JSON.stringify({ model: config.model, messages, tools: TOOLS, tool_choice: 'auto' }),
signal,
});

• "" + + "" + + "" +

临时解决方案

目前使用本地代理来添加所需的请求头，然后转发到上游 API。这虽然可行，但增加了一个额外的进程和网络跳转。

环境信息

• Evolver: 2.0.37
• OS: Windows 11
• Node: v24.18.1
• API 提供商: OpenCode Go ( + "" + https://opencode.ai/zen/go/v1 + "" + )

---

Problem

The built-in llm runner (llmRunner.js) hardcodes HTTP headers in the complete() function:

• "" + + "" + + "javascript
  headers: { 'content-type': 'application/json', authorization: \Bearer \ },
• "" + + "" + + "" +

Some OpenAI-compatible API providers require additional headers for routing/authentication. For example, OpenCode Go requires + "" + x-opencode-session + "" + for request routing:

• "" + + "" + + "json
  {
  "error": {
  "type": "MissingSessionID",
  "message": "Request is missing x-opencode-session and cannot be routed efficiently."
  }
  }
• "" + + "" + + "" +

Without this header, the API returns 400 Bad Request.

Proposed Solution

Add support for custom headers via environment variable + "" + EVOLVER_LLM_HEADERS + "" + :

Simple format (single header):

• "" + + "" + + "�ash
  export EVOLVER_LLM_HEADERS='x-opencode-session:evolver-autoexec-12345'
• "" + + "" + + "" +

JSON format (multiple headers):

• "" + + "" + + "�ash
  export EVOLVER_LLM_HEADERS='{"x-opencode-session":"evolver-12345","user-agent":"my-agent/1.0"}'
• "" + + "" + + "" +

Implementation

In + "" +
eadLlmRunnerConfig() + "" + , parse + "" + EVOLVER_LLM_HEADERS + "" + and merge into request headers:

• "" + + "" + + "javascript
  // readLlmRunnerConfig
  const rawHeaders = env['EVOLVER_LLM_HEADERS'] ?? '';
  const extraHeaders = rawHeaders.startsWith('{')
  ? JSON.parse(rawHeaders)
  : Object.fromEntries(rawHeaders.split(',').map(h => h.split(':').map(s => s.trim())));

return { apiKey, baseUrl, model, maxTurns, maxFileBytes, extraHeaders };

// complete()
const response = await fetchFn(\/chat/completions, {
method: 'POST',
headers: {
'content-type': 'application/json',
authorization: \Bearer \,
...config.extraHeaders,
},
body: JSON.stringify({ model: config.model, messages, tools: TOOLS, tool_choice: 'auto' }),
signal,
});

• "" + + "" + + "" +

Workaround

Currently using a local proxy that adds the required header before forwarding to the upstream API. This works but adds an extra process and network hop.

Environment

• Evolver: 2.0.37
• OS: Windows 11
• Node: v24.18.1
• API Provider: OpenCode Go ( + "" + https://opencode.ai/zen/go/v1 + "" + )