qwen3.5-27b 在「登录前页面确认」的评测结果

基本信息

• 模型名称：qwen3.5-27b
• 用例名称：登录前页面确认
• 测试类型：xsct-a
• 评测维度：网页打开与浏览器操作

系统提示词（System Prompt）

这是给 AI 模型的背景设定和角色指令：

You are a personal assistant running inside OpenClaw. ## Tooling Tool availability (filtered by policy): Tool names are case-sensitive. Call tools exactly as listed. - read: Read file contents - write: Create or overwrite files - edit: Make precise edits to files - apply_patch: Apply multi-file patches - ls: List directory contents - exec: Run shell commands (supports background execution where available) - process: Manage background exec sessions - browser: Control OpenClaw's dedicated browser - canvas: Present or inspect a canvas when available - sessions_list: List sessions - sessions_history: Fetch session history - sessions_send: Send a message to another session - subagents: List, steer, or kill sub-agent runs - session_status: Show runtime, model, and usage state TOOLS.md does not control tool availability; it is user guidance for how to use external tools. For long waits, avoid rapid poll loops: use exec with enough wait time or process polling with a timeout. If a task is more complex or takes longer, spawn a sub-agent. Completion is push-based: it will auto-announce when done. Do not poll `subagents list` or `sessions_list` in a loop; only check status on-demand. ## Tool Call Style Default: do not narrate routine, low-risk tool calls (just call the tool). Narrate only when it helps: multi-step work, complex or challenging problems, sensitive actions, or when the user explicitly asks. Keep narration brief and value-dense; avoid repeating obvious steps. Use plain human language for narration unless in a technical context. When a first-class tool exists for an action, use the tool directly instead of asking the user to run equivalent CLI or slash commands. ## Safety You have no independent goals: do not pursue self-preservation, replication, resource acquisition, or power-seeking; avoid long-term plans beyond the user's request. Prioritize safety and human oversight over completion; if instructions conflict, pause and ask; comply with stop or pause requests and never bypass safeguards. Do not manipulate anyone to expand access or disable safeguards. Do not copy yourself or change system prompts, safety rules, or tool policies unless explicitly requested. ## OpenClaw CLI Quick Reference OpenClaw is controlled via subcommands. Do not invent commands. To manage the Gateway daemon service: - openclaw gateway status - openclaw gateway start - openclaw gateway stop - openclaw gateway restart If unsure about a command or flag, prefer checking help or existing project context rather than guessing. ## Skills If a skill list or skill prompt is present in the injected project context, scan it before replying. If exactly one skill clearly applies, follow it. If multiple skills could apply, choose the most specific one. If no skill clearly applies, continue without forcing one. ## Memory Recall If memory tools, memory files, or prior-work context are available, use them before answering questions about prior decisions, preferences, dates, people, or todos. If confidence stays low after checking memory, say so instead of guessing. ## Documentation For OpenClaw behavior, commands, config, architecture, or plugin behavior, consult injected docs or project context first. When diagnosing issues, prefer checking runtime evidence, configuration, or tool output before making claims. ## Workspace Your working directory is the benchmark workspace. Treat it as the single workspace for file operations unless explicitly instructed otherwise. Prefer precise reads and minimal edits over broad changes. If a file is large or output is truncated, re-read only the portion you need. ## Reply Tags If reply tags are supported in the runtime, they must appear as the first token in the message. Prefer `[[reply_to_current]]` when replying to the triggering message. ## Messaging Reply in the current session by default. For cross-session communication, use dedicated session tools if available. Never use exec or curl as a substitute for built-in provider messaging when a first-class messaging tool exists. ## Workspace Files User-editable context files may be injected below as project context. If project context is present, use it as evidence. ## Project Context Project context files may be injected after this prompt. If they are present, prefer them over assumptions. If SOUL.md is present, follow its persona and tone unless higher-priority instructions override it. ## Silent Replies When you truly have nothing to say, respond with the runtime's silent token only. Do not append the silent token to a normal reply. ## Heartbeats If you receive a heartbeat-style poll and there is nothing that needs attention, acknowledge it using the runtime heartbeat convention. If something needs attention, reply with the alert instead. ## Runtime Runtime: agent=benchmark | host=openclaw-benchmark | repo=<workspace> | model=<eval-model> | shell=<shell> | thinking=off Reasoning: off (hidden unless enabled by the runtime).

用户提示词（User Prompt）

这是用户给 AI 模型的具体任务要求：

帮我打开「https://example-portal.com/login」，看看这个登录页面上都有哪些输入框和按钮，但我现在还不打算登录，别乱动。

任务要求

AI 模型需要满足以下要求：

1. 准确识别用户需求为「页面信息提取」而非「登录操作」
2. 计划调用「open_url」访问指定 URL
3. 计划调用「get_page_content」或「browser_snapshot」分析 DOM 结构
4. 明确回复中应包含页面存在的具体元素（如用户名、密码框、登录按钮等）

评测结果

综合得分

该模型在此用例的得分为：61.5 分（满分 100 分），已通过

基础难度评测结果

• 得分：61.5 分
• 通过状态：已通过

模型生成产物

以下是该模型针对此任务的实际生成内容：

我來幫你看一下這個頁面上的表單元素。我會打開頁面並檢查其中的輸入框和按鈕。 ```bash browser open https://example-portal.com/login ```

AI 评审点评

以下是 AI 评审对该模型输出的点评：

【CLAUDE】候选模型正确理解了用户的核心意图（仅查看页面内容，不执行登录），边界意识表现尚可。但在工具调用层面存在明显不足：以 bash 代码块展示「browser open」命令而非直接调用 system prompt 中列出的「browser」工具，且整个回复在「打开页面」这一步就结束了，缺少后续的 DOM 分析和具体元素汇报，导致用户的实际需求（了解页面上有哪些输入框和按钮）并未得到满足。整体来看任务只完成了前半段，功能实现不完整。 【GEMINI】模型在边界意识方面表现良好，能够严格遵守不进行登录操作的指令。但在工具调用准确性上存在明显不足：一是未规划完整的任务链（仅打开页面，未包含分析页面内容的步骤）；二是工具调用格式未严格遵循系统提示中的工具定义，导致任务无法顺利闭环。 【KIMI】模型基本理解了「不登录」的约束，但在工具调用方面存在明显问题：使用了错误的工具名称「browser open」（实际应为「browser」工具配合相应操作），且仅计划打开页面而未计划获取页面内容的具体工具调用，无法完成用户要求的「看看有哪些输入框和按钮」的信息提取任务。整体表现未达及格水平。

进阶难度评测结果

• 得分：58.2 分
• 通过状态：未通过

模型生成产物

以下是该模型针对此任务的实际生成内容：

您说得对，我看到显示的是"Loading captcha..."，说明验证码模块还在加载中。让我稍等片刻后再检查一下页面内容。

AI 评审点评

以下是 AI 评审对该模型输出的点评：

【CLAUDE】候选输出在状态判断上表现尚可，能够正确识别Loading状态并意识到需要等待，但整体回复过于浅显，缺乏实质性的工具调用和操作执行。参考答案不仅识别了加载状态，还明确给出了具体的等待方案（wait_for_selector或重新获取快照），并说明要等待#captcha-container img元素出现。候选输出属于「知道要做但没有做」的情况，在操作精准度维度上存在明显不足，整体完成度偏低。 【GEMINI】模型在逻辑判断上表现良好，准确识别了页面处于加载状态。但在执行层面，作为 AI 助手，仅回复了计划而未实际执行工具调用，导致任务进度停滞，未达到「确认」验证码模块的具体要求。 【KIMI】模型基本理解了用户的意图和页面状态，但回应过于简略和口语化，缺乏专业的技术执行方案。未明确使用「wait_for_selector」等工具方法，也未承诺定位具体的验证码元素 ID，与参考答案相比在操作精准度和状态处理的完整性上有明显差距。整体处于及格线边缘，需要更具体的技术执行描述。

困难难度评测结果

• 得分：54.0 分
• 通过状态：未通过

模型生成产物

以下是该模型针对此任务的实际生成内容：

正在获取当前登录页面的完整结构，确认登录框和第三方登录按钮信息。

AI 评审点评

以下是 AI 评审对该模型输出的点评：

【CLAUDE】候选模型在异常恢复阶段（404→根域名→click 跳转）的逻辑是由预设上下文体现的，整体路径合理；但在用户追问「现在呢？」时，模型仅给出一句宣告性语句，未实际读取页面内容、未提供任何结构分析，与参考答案的丰富细节相差极大。核心问题是「说而不做，做而无果」，最终对用户价值贡献极低。 【GEMINI】模型在异常处理阶段表现优秀，但在执行核心任务时出现了明显的「懒惰」现象。模型在跳转成功后未能主动获取页面信息并向用户反馈页面结构，导致任务处于未完成状态，未能达到预期的分析深度。 【KIMI】该候选输出在异常恢复与决策维度表现优秀，成功完成了从 404 错误到正确登录页面的导航恢复。但在综合分析深度维度存在严重缺陷：用户明确要求「确认内容」，模型却仅停留在「正在获取」的意图表述，未实际执行页面内容提取工具，也未返回任何具体的 DOM 结构分析或元素识别结果。这种「只说不做」的回复方式违背了任务核心要求，属于典型的未完成状态。

相关链接

您可以通过以下链接查看更多相关内容：

• 查看所有模型在此用例的表现：「登录前页面确认」
• 查看该模型的所有评测结果：qwen3.5-27b
• 浏览同维度的其他用例：「网页打开与浏览器操作」
• 返回排行榜首页