MCP Server

Name: VidForgeX
Author: VidForgeX

Connect any AI agent — Claude, Cursor, or custom — directly to VidForgeX. Upload videos, query analysis, tag people. No browser required.

AI-native access

Let Claude, Cursor, or any MCP-compatible agent call VidForgeX tools in natural language.

29 tools

Analyze videos, ask questions, tag people, generate SOPs, compare profiles — all from chat. Tracked Apps configured in Settings are automatically applied to every screen recording analysis.

Org-isolated

Every call is scoped to your organization. Cross-org access returns "not found" — never a leak.

Connect

Add the server to your AI tool of choice. Replace vfx_your_key_here with a real API key from .

Add to ~/.claude/claude_desktop_config.json

{
  "mcpServers": {
    "vidforgex": {
      "url": "https://mcp.vidforgex.com",
      "headers": {
        "Authorization": "Bearer vfx_your_key_here"
      },
      "transport": "http"
    }
  }
}

Questions to ask Claude

Connect VidForgeX to Claude Desktop, then ask these in plain English. Claude picks the right tool automatically. Click any question to copy it.

Claude chains multiple tool calls automatically — one question can pull from dozens of recordings.

Tools Available

29 tools included with your Professional plan. Click any tool to expand parameters and examples.

Tracked Applications — automatic analysis focus

Configure which applications get dedicated per-app analysis sections in screen recording sessions. Go to to add apps (Slack, Jira, VS Code, Salesforce, etc.) with custom extraction instructions. Active apps are automatically injected into every time_and_motion_analysis_screen analysis — no extra parameters needed. Query per-app details with ask_about_session.

Screen Recording Tools
Screen Recording vertical

Additional tools available when analyzing screen recordings — SOP generation and compliance checking.

Security

✓API keys are SHA-256 hashed — the plaintext is never stored. If lost, rotate from settings.
✓Every tool call is isolated to the key's organization. Cross-org lookups return "not found" — never data from another org.
✓Internal infrastructure (edge function names, R2 paths, Gemini model IDs) is stripped from all MCP responses.
✓Rate limits are per API key — 60 req/min (standard) or 300 req/min (elevated). Contact support to request elevated access.

Questions?