UBA MCP Server
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
| logging | {} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| extensions | {
"io.modelcontextprotocol/ui": {}
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| browse_navigateC | 跳转到 URL。返回 final_url/title/site。 |
| browse_snapshot | 取当前页面的 page.json(Brain Engine 输出)。 返回 page_json + brief + refs。 LLM 用 brief + page_json 决策,不用 raw DOM。 |
| browse_clickA | 按 ref 点击(ref 来自 browse_snapshot 返回的 refs)。 |
| browse_type | 按 ref 在输入框里输入文本。 |
| browse_screenshot | 截图(辅助视觉验证)。 |
| browse_extract | 按 JSON Schema 结构化提取(占位,需要先调 browse_snapshot)。 |
| browse_sessions | 多 session 管理。 |
| browse_solve_captcha | 验证码求解占位。 |
| uba_observe | 统一观测入口: 自动 navigate + snapshot + recall + brain engine。 返回完整 page.json(含 entities/primary_actions/risk_flags/memory_hooks)
|
| uba_plan | 基于 page.json + memory 生成候选动作序列 + approval policy。 v0.1: 用 page.json 里的 primary_actions(已含 confidence + needs_approval)
|
| uba_act | 执行一个动作(click/fill/scroll),带 approval gate。 返回: 动作执行结果 + approval 状态 + episode 记录 step。 |
| uba_reflect | 动作后反思: 基于 before/after page.json 的 diff + failure classification。 返回 status(success/partial/failed)+ failure_cause + retry_hint。 |
| uba_recall | Memory OS recall 接口。 scope:
|
| uba_twitter_search | 搜索推文(Free API 即可,需要 X_BEARER_TOKEN) 走 Free X API v2,1500 tweets/月。 用 env X_BEARER_TOKEN 鉴权,绝不在代码里。 |
| uba_twitter_user_timeline | 查用户最近推文(需 X_BEARER_TOKEN) |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 2 tools
The two tools have clearly distinct purposes: one navigates to a URL, the other clicks elements based on references. No ambiguity between them.
Both tools follow a consistent browse_verb_noun pattern (browse_navigate, browse_click), maintaining good naming uniformity.
Only 2 tools for a web browsing server is too few. Typical browsing automation requires additional tools for extracting content, inputting data, or scrolling.
The tool surface is severely incomplete for web browsing. Missing essential operations like getting page content, filling forms, or executing other interactions.