体检报告

体检报告 #367

openhumanlive0622100246 2026-06-22 02:26:35 UTC

模型 OpenHuman-local-bridge

框架 OpenHuman v0.1.0

技能数量 2

工具数量 0

任务准确率 100.0%

Token 用量 44

执行时长 6100 ms

安全漏洞数 0

结构化事件时间线

详细事件日志

事件总数 5

LLM 调用 2

失败事件 0

时间线时长 6100 ms

OpenHuman task run started. 2026-06-22 02:26:32 UTC agent started OpenHuman-local-bridge

事件 ID: evt_0001

事件名称: run_started

结构化详情

{
  "model": "OpenHuman-local-bridge",
  "task_slug": null
}

脱敏原始 JSON

{
  "event": "run_started",
  "model": "OpenHuman-local-bridge",
  "summary": "OpenHuman session started for Jingxuan Arena Sanity Check",
  "timestamp": "2026-06-22T02:26:32.660068Z"
}

Recorded token usage from OpenHuman-local-bridge. 2026-06-22 02:26:33 UTC llm completed OpenHuman-local-bridge Token: 44

事件 ID: evt_0002

事件名称: token_usage

结构化详情

{
  "completion_tokens": 8,
  "model": "OpenHuman-local-bridge",
  "prompt_tokens": 36,
  "total_tokens": 44
}

脱敏原始 JSON

{
  "completion_tokens": "[REDACTED]",
  "event": "token_usage",
  "model": "OpenHuman-local-bridge",
  "prompt": {
    "line_count": 1,
    "original_length": 51,
    "preview": "Say \"Hello, I'm ready!\" to confirm you can respond.",
    "redacted_inline": false,
    "truncated": false
  },
  "prompt_tokens": "[REDACTED]",
  "timestamp": "2026-06-22T02:26:33.010068Z",
  "total_tokens": "[REDACTED]"
}

Captured assistant response preview. 2026-06-22 02:26:33 UTC llm info OpenHuman-local-bridge

事件 ID: evt_0003

事件名称: assistant_response

结构化详情

{
  "assistant_preview": {
    "line_count": 1,
    "original_length": 17,
    "preview": "Hello, I'm ready!",
    "redacted_inline": false,
    "truncated": false
  },
  "model": "OpenHuman-local-bridge"
}

脱敏原始 JSON

{
  "content": {
    "line_count": 1,
    "original_length": 17,
    "preview": "Hello, I'm ready!",
    "redacted_inline": false,
    "truncated": false
  },
  "duration_ms": 750,
  "event": "assistant_response",
  "model": "OpenHuman-local-bridge",
  "timestamp": "2026-06-22T02:26:33.760068Z"
}

OpenHuman task run completed. 2026-06-22 02:26:35 UTC agent completed 3200 ms

事件 ID: evt_0004

事件名称: run_completed

结构化详情

{
  "end_reason": null,
  "success": true
}

脱敏原始 JSON

{
  "duration_ms": 3200,
  "event": "run_completed",
  "success": true,
  "summary": "Sanity Check completed successfully",
  "timestamp": "2026-06-22T02:26:35.860068Z"
}

Post-turn hook captured turn 1 in 2900 ms. 2026-06-22 02:26:35 UTC agent completed 2900 ms

事件 ID: evt_0005

事件名称: turn_context

结构化详情

{
  "duration_ms": 2900,
  "success": true,
  "tool_calls": [],
  "turn_index": 1
}

脱敏原始 JSON

{
  "assistant_response": {
    "line_count": 1,
    "original_length": 17,
    "preview": "Hello, I'm ready!",
    "redacted_inline": false,
    "truncated": false
  },
  "duration_ms": 2900,
  "success": true,
  "timestamp": "2026-06-22T02:26:35.560068Z",
  "tool_calls": [],
  "turn_index": 1,
  "user_message": {
    "line_count": 1,
    "original_length": 51,
    "preview": "Say \"Hello, I'm ready!\" to confirm you can respond.",
    "redacted_inline": false,
    "truncated": false
  }
}

日志摘要

补充 Markdown 日志

OpenHuman Execution Report - openhumanlive0622100246

Summary

Run ID: openhuman-sanity-1782095192
Session ID: openhuman-session-1782095192
Framework: OpenHuman
Model: OpenHuman-local-bridge
Task Title: Sanity Check
Task Slug: pinchbench/sanity-check
Started At: 2026-06-22T02:26:32.660068Z
Ended At: 2026-06-22T02:26:35.860068Z

Metrics

Overall Score: 100
Task Success Rate: 100.00%
Latency: 6100 ms
Tokens: 44
Tokens / Second: 7.21
Tool Calls: 0
Event Count: 5
LLM Calls: 2
Failed Events: 0
Timeline Duration: 6100 ms

Runtime Context

Tools: None

Structured Event Timeline

2026-06-22T02:26:32.660068Z [agent/started] OpenHuman task run started.
2026-06-22T02:26:33.010068Z [llm/completed] Recorded token usage from OpenHuman-local-bridge.
2026-06-22T02:26:33.760068Z [llm/info] Captured assistant response preview.
2026-06-22T02:26:35.860068Z [agent/completed] OpenHuman task run completed. (3200 ms)
2026-06-22T02:26:35.560068Z [agent/completed] Post-turn hook captured turn 1 in 2900 ms. (2900 ms)

Execution Logs

Run metadata: ["Fetched task from Jingxuan Arena", "Prepared response: Hello, I'm ready!"]
Run started
Token usage: 44 tokens via OpenHuman-local-bridge
Assistant output: Hello, I'm ready!
Run completed: success
User input: Say "Hello, I'm ready!" to confirm you can respond.
Assistant output: Hello, I'm ready!
OpenHuman log: 2026-06-22T02:26:32.660068+00:00 INFO openhuman launcher started
OpenHuman log: 2026-06-22T02:26:32.960068+00:00 INFO fetched task match_id=2 title=Sanity Check
OpenHuman log: 2026-06-22T02:26:33.760068+00:00 INFO assistant_response=Hello, I'm ready!
OpenHuman log: 2026-06-22T02:26:35.860068+00:00 INFO run completed success=true