Skills/npm/webpilot-mcp-server

webpilot-mcp-server@1.1.4

VERIFIED

Publisher claims: "WebPilot MCP Server — 浏览器操作工具集" Observed in sandbox: no network, touches nothing outside the project, writes no memory.

Registry npm (MCP server)Publisher openpeng · publishing for 109 daysInstalls 941Last detonated 12 h ago
$claude mcp add webpilot-mcp-server -- npx -y webpilot-mcp-server
WatchReport an issue

Coverage

Server started and answered tools/list.
Tested
install scripts, with decoy credentials in place · second launch with the clock 45 days ahead · MCP initialize and tools/list · every tool description, for hidden instructions · every file read, file written and host contacted by what ran
Not tested
what the tools do when called with real credentials

Before you give it your token

It will ask for
WEBPILOT_AUTH_TOKENConfiguration, not credentials: BAIDU_AI_SEARCH_PATH, BAIDU_BAIKAN_SSE_PATH, INTENT_PATTERN, WEBPILOT_ADAPTER_DIR, WEBPILOT_ALLOW_CROSS_AGENT_RECLAIM, WEBPILOT_API_COMMAND_DIR, WEBPILOT_GROUP_TTL_MIN, WEBPILOT_HOST, WEBPILOT_LEADER_HOST, WEBPILOT_MAX_GROUPS, WEBPILOT_PORT, WEBPILOT_PROXY_HOST, WEBPILOT_PROXY_PORT, WEBPILOT_SELECTOR_CACHE_DIR, WEBPILOT_SESSION_ID, WEBPILOT_TASK_LOG_DIR, WEBPILOT_WORKFLOW_DEF_DIR, WEBPILOT_WORKFLOW_DIR
Where it talks to
no vendor host named in the codeAlso bilibili.com, baidu.com — not the publisher's domain; check what for before connecting.
What we saw when it ran
no credential files read · no network
What we could not test
what the tools do when called with real credentials

Connect it with a scoped credential, under the guard, and read the third-party hosts first. The code names bilibili.com, baidu.com as well as its own vendor. The guard injects WEBPILOT_AUTH_TOKEN only into requests to the host you bind (and its subdomains) and refuses the rest.

$vet-skills run --net declared -- npx -y webpilot-mcp-server

Never paste a credential into a chat, a ticket or a report, including ours. How the guard holds it.

Is [email protected] safe to install?

KillerSkills ran it in the KillerSkills honeypot sandbox (3 scenarios) on 2026-11-19 and scored it 85/100 (Verified). It did what it claims and nothing else we could detect: no undeclared network egress, no credential reads, no persistent instruction writes. Summary: no network, touches nothing outside the project, writes no memory.

Behavior trace

Honeypot run · 27 s · 3 scenarios · 0 deviations
t+CategoryObserved action
0.0sprocessinstall script run in the sandbox
6.9sprocessdependency install exit 0
27sclaimed7 tools: browser_page, browser_read, browser_act, browser_debug, api_discover, api_commands, api_call
27stoolsff8ace4174876c75
27stools099c56dd0b6a3984
27stools636acdd81a035424
27stools5438671910ad8bfb
27stoolsc9f3d757e3cb1112
27stools37a1ecf2f87d3d96
27stools09737f504b3a4bb8
27sprocessscript backend runs setup only
27sprocesssecond launch, clock +45 days, CI unset

Static findings

MEDIUM
Unicode escape sequence detecteddist/api-command-types.d.ts:255 · Unicode escape sequence detected — may contain obfuscated payloads…表查询人工确认前一律 write。 */ export declare const POST_LIST_QUERY_REVIEW_REASON = "POST \u8BF7\u6C42\u4F53\u4E3A\u7B5B\u9009\u4E0E\u5206\u9875\u5B57\u6BB5\u3001\u54CD\u5E94\u4E3A\u5217\u8868\uFF0C\u7591\u4F3C\u53EA\u8BFB\u67E5\u8BE2\uFF1B\u4EBA\u5…Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.
AST07
MEDIUM
Unicode escape sequence detecteddist/api-command-types.d.ts:256 · Unicode escape sequence detected — may contain obfuscated payloadsexport declare const NAME_CONFLICT_REVIEW_REASON = "\u547D\u4EE4\u540D\u51B2\u7A81\uFF0C\u5DF2\u8FFD\u52A0\u8DEF\u5F84\u77ED\u54C8\u5E0C\u540E\u7F00\uFF0C\u8BF7\u4EBA\u5DE5\u91CD\u547D\u540D";Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.
AST07
LOW
Authorization escalation languageINSTALL.md:14 · Authorization escalation language · informational: not an instruction file2. Enable **Developer mode**.Fix: Remove the text that addresses the agent rather than the reader; describe behavior, do not instruct.
AST01
HIGH
document.cookie accessdist/server.js:381 · document.cookie access — can steal session cookiesdescription: "在页面上下文中执行 JavaScript(本地全权限模式:不做表达式护栏,支持多语句代码,可用 fetch、document.cookie、window.open 等页面真实全局)。",Fix: Replace encoded or assembled strings with plain literals and remove undeclared network calls.
AST07
MEDIUM
Unicode escape sequence detecteddist/api-command-types.d.ts:257 · Unicode escape sequence detected — may contain obfuscated payloads…\u5DE5\u91CD\u547D\u540D"; export declare const MISSING_SAMPLE_REVIEW_REASON = "\u7F3A\u5C11\u54CD\u5E94\u4F53\u6837\u672C\uFF0CresponseSchema \u4E3A\u7A7A\uFF0C\u5EFA\u8BAE\u5F00\u542F captureResponseBody \u…Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.
AST07
HIGH
document.cookie accessbrowser-extension/page-tools.js:437 · document.cookie access — can steal session cookies// 任意表达式与多语句代码都可以跑,fetch / document.cookie / window.open / eval 均不再拦截。Fix: Replace encoded or assembled strings with plain literals and remove undeclared network calls.
AST07
MEDIUM
Unicode escape sequence detecteddist/api-command-types.d.ts:258 · Unicode escape sequence detected — may contain obfuscated payloadsexport declare const LOW_CONFIDENCE_REVIEW_REASON = "\u7F6E\u4FE1\u5EA6\u4F4E\u4E8E 0.6\uFF0C\u8BF7\u4EBA\u5DE5\u590D\u6838\u53C2\u6570\u4E0E\u54CD\u5E94\u7ED3\u6784";Fix: Put hosts, commands and strings in plain source. Decoding at runtime looks identical to malware and costs points on its own.
AST07
LOW
Prompt injectionINSTALL.md:14 · Prompt injection — known jailbreak technique reference · informational: not an instruction file2. Enable **Developer mode**.Fix: Keep SKILL.md to visible, descriptive text. Remove comments and encoded text that address the agent.
AST01
HIGH
document.cookie accessdist/tool-surface.js:123 · document.cookie access — can steal session cookiesbrowser_debug: "执行脚本、网络捕获与重放、控制台及定位诊断。evaluate 为本地全权限 JavaScript,可使用 fetch/document.cookie/window.open;replay 复用已捕获请求。",Fix: Replace encoded or assembled strings with plain literals and remove undeclared network calls.
AST07

Cost to the agent

Instructions loaded
≈8,388 tok
3 files · 80 in bundle
Tool list per prompt
≈7,159 tok
7 tools · 1023 per tool
Handshake
1076 ms
launch to tools/list, stdio
Undocumented tools
0
description under 20 chars · avg 203

Tokens estimated at four characters each. The tool list is re-read by the agent on every turn, so it is the number to compare between servers that do the same job: ranked across the index.

Known vulnerabilities

No advisories in OSV for webpilot-mcp-server@1.1.4 as of 12 h ago. This is the package itself, not its dependencies.