$ timeahead.in
/servers/blog/about
/ servers/npm/fsb-mcp-server
npm

fsb-mcp-server

FSB Browser Automation MCP Server

★ 180 stars↓ 37/wkupdated 1d agogithub ↗
☆ Watch this server📋 Install guide⚡ Compare⚑ Claim listing
78fair
▣ Overview
FSBVersionManifest V3LicenseStarsForksIssuesLast Commit

What it does

FSB is a Chrome extension for browser automation through natural language descriptions. Rather than relying on visual page analysis, it reads the live DOM structure, identifies clickable elements and forms, builds an execution plan, and performs browser actions. When set up as an MCP server, Claude Code and other MCP clients can drive the browser by describing tasks in plain English—no browser automation code required. The extension maintains session state, verifies action outcomes, and reports progress and logs.

Who it's for

QA engineers automating regression test flows, researchers extracting and comparing information across multiple pages, and developers building AI-powered workflows that require browsing without writing browser automation scripts.

Common use cases

  • Run end-to-end QA test flows described in plain language
  • Extract information from multiple pages and summarize findings
  • Fill and submit forms across different sites
  • Verify page state changes and collect action logs for regression testing
  • Automate multi-step workflows that require conditional branching based on page state

Setup pitfalls

  • Marked high-risk due to filesystem access, network requests, and active browser control—behavior can be unpredictable on complex or sensitive sites; test on non-critical pages first
  • Requires both the Chrome extension (installed from Chrome Web Store) and the npx fsb-mcp-server installation; each must be set up and the client must be restarted separately
  • Each browser profile executes one automation task at a time; parallel automation requires separate browser instances or profiles
  • If using the extension side panel directly (not through MCP), provide an API key for an AI provider (xAI, Gemini, OpenAI, Anthropic, etc.); MCP clients can omit this requirement
▣ Score BreakdownMCPScore = Σ(raw × weight)
DimensionRawWeighted
Security
35%
100
35.0
Freshness
25%
85
21.3
Adoption
20%
34
6.8
Quality
10%
100
10.0
Trust
10%
50
5.0
Total
78.0
⚿ Capabilities & Risk Explainer
fs readfs writenetworkexecevalsecrets
◆ Risk level: high· 15 tools · auth: API key
fs read + fs write + network + exec + eval + secrets active — can execute code, access credentials, and make external network calls.
Tool nameDescriptionDestructive?
back✓ no
list_sessions✓ no
get_session_detail✓ no
get_logs✓ no
search_memory✓ no
+10 more tools
get_memory_stats✓ no
run_task⚠⚠ yes
stop_task✓ no
get_task_status✓ no
list_credentials✓ no
fill_credential✓ no
list_payment_methods✓ no
use_payment_method✓ no
start_visual_session✓ no
end_visual_session✓ no
⚙ Install config
Claude Desktop · Cursor · Windsurf · VS Code (Copilot) · Claude Code
add to your MCP client config:
{
  "mcpServers": {
    "fsb": {
      "command": "npx",
      "args": [
        "-y",
        "fsb-mcp-server"
      ]
    }
  }
}
📈 Score historylast 44 snapshots
4/30/20267/1/2026 · 44 snapshots
⚙ Maintenance health
63/ 100 · is this project alive?
contributors (1y)2
top contributor share82%
releases (1y)23
last release8d ago
ci✓ passing
⛁ Raw data
weekly downloads37
github stars180
forks12
open issues0
license✓ present
readme length33891 chars
last publish1d ago
last commit1d ago
last updated6d ago
install verified✗ fail · 47d ago
owner of this server? claim your listing to get a verified badgeclaim →
🔔 Score drop alerts
get notified by email when this server's score drops 5+ points