datawhalechina/hello-agents · error · HTTPException

重置失败: {str(e)}

Error message

重置失败: {str(e)}

What it means

Catch-all handler for POST /config/reset: any exception raised while clearing sessions, memory, or global config is re-raised as HTTP 500 with '重置失败: <original message>'. The detail string is the only diagnostic surface — the original traceback goes to server logs. Causes range from locked/missing files to LLM failures during re-summarization, so read the embedded message to classify.

Source

Thrown at Co-creation-projects/tino-chen-HelloClaw/src/api/config.py:167

        )

        # 如果清除了会话，也要清除 Agent 内存中的历史记录
        if reset_sessions:
            agent = get_agent()
            if agent:
                agent.clear_all_history()

        messages = ["配置文件已重置"]
        if reset_sessions:
            messages.append("会话已清除")
        if reset_memory:
            messages.append("每日记忆已清除")
        if reset_global_config:
            messages.append("全局配置已重置")

        return {"status": "success", "message": "，".join(messages)}
    except Exception as e:
        raise HTTPException(status_code=500, detail=f"重置失败: {str(e)}")


@router.get("/agent/info", response_model=AgentInfo)
async def get_agent_info(ws: WorkspaceManager = Depends(get_workspace)):
    """获取助手信息（包括名字）

    每次都重新读取 IDENTITY.md 以获取最新的名字
    """
    # 从 IDENTITY.md 读取最新的名字
    identity = ws.load_config("IDENTITY")
    name = "HelloClaw"  # 默认名字

    if identity:
        import re
        # 匹配格式: - **名称：** xxx 或 - **名称:** xxx
        match = re.search(r'\*\*名称[：:]\*\*\s*(.+?)(?:\n|$)', identity)
        if match:
            name = match.group(1).strip()

View on GitHub (pinned to 606a07d341)

Solutions

  1. Read the embedded exception text in the 500 detail — it names the real failing operation
  2. Check server stdout/logs for the full traceback at the moment of the request
  3. Fix filesystem permissions on the workspace tree (chown -R or chmod) and retry
  4. Narrow the reset scope (only one flag at a time) to isolate which branch throws
  5. Restart the server so the global Agent is initialized before resetting sessions
Defensive patterns

Strategy: try-catch

Try / catch

try {
  const res = await fetch('/api/config/reset', {method:'POST', body: JSON.stringify(payload)});
  if (!res.ok) {
    const {detail} = await res.json();
    report(detail); // detail embeds the real exception message
  }
} catch (e) { /* network-level failure */ }

Prevention

When it happens

Trigger: POST /api/config/reset with reset_sessions=true while history files are missing or unreadable; reset_global_config=true when config.json path is not writable; agent global not initialized when a reset branch calls into agent.clear_all_history(); permission errors when the workspace is owned by another user.

Common situations: Running the server from a different user/container than the one that created ~/.helloclaw; workspace on a read-only mount; concurrent reset during an active chat writing to the same session files; first-boot reset before agent initialization completed.

Related errors

AI-assisted analysis of datawhalechina/hello-agents@606a07d341 (2026-08-14). Data as JSON: /api/errors/88fbd90bd2377e48. Report an issue: GitHub.