For AI agents — MCPBundles Directory API
For MCP client apps (human connect):
MCP connect URL: https://mcp.mcpbundles.com/bundle/cursor
Skill page: https://www.mcpbundles.com/skills/cursor
Tools: 33
Live probe refreshed Nov 21, 2026 · Endpoint host mcp.mcpbundles.com
What you can do with Cursor
Built for
Engineering leads, platform teams, and developers automating Cursor Cloud Agents and team insights
MCPBundles publishes this directory for MCP discovery. Except where we host or operate an offering, third-party services run under their own terms. Product and company names on this page are used in a descriptive, identifying way (including under nominative fair use where applicable); they remain the property of their owners. Nothing here grants you rights in those marks, and nothing here is an offer to sell a third party's services. Terms
Cursor
Add to your AI client, then sign in once when you connect.
Launch Background Agents on GitHub repos, follow runs with follow-ups, and pull team usage, admin, and AI-code analytics from one Cursor API key.
This listing identifies a third-party integration. MCPBundles is not affiliated with, sponsored by, or endorsed by the service or its operator unless stated on this page.
List Cursor models and repositories I can use, then launch a Background Agent on https://github.com/my-org/my-repo with a prompt to add tests for the auth module and open a branch.
Check running agents
Operational monitoring without the Cursor dashboard.
Fetch Cursor team audit logs for the last 7 days filtered to login and settings_changed events and list anything unusual.
Context to know first
Do I need different API keys for agents vs admin analytics?
No separate MCPBundles connections—one Cursor API key field. Cursor decides what each key can call: Background Agents need on-demand usage on the account; team admin and analytics endpoints need a team key with admin access from Cursor team settings.
Why did launch fail with on-demand usage?
The Cursor account that owns the API key must enable on-demand usage for Background Agents in Cursor Dashboard → Settings. Rotating the key alone does not remove that requirement.
How many tools are included?
The bundle exposes the full Cursor API surface on MCPBundles—Cloud Agents lifecycle, generic fetch/delete helpers, team admin, analytics, and AI-code tracking—within one connected server.
AI Skill
SKILL.md
Domain knowledge for Cursor — workflow patterns, data models, and gotchas for your AI agent.
Cursor
Cursor Cloud Agents API for programmatically launching and managing AI-powered coding agents that autonomously work on your GitHub repositories. Full agent lifecycle management: create agents with prompts, add follow-ups, monitor status, and manage branches/PRs.
Before launching a Background Agent
Launch calls fail with HTTP 400 until the Cursor account that owns the API key has Background Agents billing ready:
Rotating the API key alone does not bypass this gate—the upstream error names on-demand usage explicitly.
Tell the human to fix Cursor account settings, then retry launch; do not loop launch attempts while the account is still blocked.
Repository listing is rate-limited (about one request per minute). Prefer a known https://github.com/owner/repo URL when the repo is already identified.
Data Model
Agents, Models, Key Info, Repositories — each has a listing read.
Admin, analytics, and AI-code tracking endpoints use the same saved API key (Basic auth). Cursor may require a team admin key from team settings for /teams/* and team analytics; Cloud Agents still need on-demand usage on the owning account. A 403 usually means the key lacks team scope—not a separate MCPBundles connection.
Add a follow-up instruction to an existing cloud agent while it's running. Use this to refine the agent's work, add additional requirements, or request changes without starting a new agent. The agent will build on its previous changes and apply the new instruction. Only works with RUNNING agents; cannot add follow-ups to FINISHED, FAILED, or CANCELLED agents. Returns the agent ID for tracking the continued work.
Cursor Add Followup — Add a follow-up instruction to an existing cloud agent while it's running. Use this to refine the agent's work, add additional requirements, or request changes without starting a new agent. The agent will build on its previous changes and apply the new instruction. Only works with RUNNING agents; cannot add follow-ups to FINISHED, FAILED, or CANCELLED agents. Returns the agent ID for tracking the continued work.
Cursor Admin Audit Logs — Retrieve audit log events for your team with filtering. Track team activity, security events, and configuration changes. Rate limited to 20 requests per minute per team. Date range cannot exceed 30 days.
Cursor Admin Daily Usage Data — Retrieve daily usage metrics for your team. Data is aggregated at the hourly level - we recommend polling this endpoint at most once per hour. Rate limited to 20 requests per minute per team. Date range cannot exceed 30 days.
Cursor Admin Delete Repo Blocklist — Remove a specific repository from the blocklist. Returns 204 No Content on successful deletion.
Cursor Admin Get Repo Blocklists — Retrieve all repository blocklists configured for your team. Add repositories and use patterns to prevent files or directories from being indexed or used as context.
Cursor Admin Set User Spend Limit — Set spending limits for individual team members. This allows you to control how much each user can spend on AI usage within your team. Rate limited to 60 requests per minute per team. Only integer values are accepted (no decimal amounts). Setting spendLimitDollars to null will clear/remove the limit entirely.
Cursor Admin Spending Data — Retrieve spending information for the current calendar month with search, sorting, and pagination.
Cursor Admin Team Members
Frequently Asked Questions
Do I need different API keys for agents vs admin analytics?
No separate MCPBundles connections—one Cursor API key field. Cursor decides what each key can call: Background Agents need on-demand usage on the account; team admin and analytics endpoints need a team key with admin access from Cursor team settings.
Why did launch fail with on-demand usage?
The Cursor account that owns the API key must enable on-demand usage for Background Agents in Cursor Dashboard → Settings. Rotating the key alone does not remove that requirement.
Standard connector25 free calls every 30 days, then from $19/month
$19 / month
Subject to fair usage limits
— Retrieve all team members and their details including name, email, and role.
Cursor Admin Upsert Repo Blocklists — Replace existing repository blocklists for the provided repos. This endpoint will only overwrite the patterns for the repositories provided. All other repos will be unaffected.
Cursor Admin Usage Events Data — Retrieve detailed usage events for your team with full filtering, search, and pagination options. Provides granular insights into individual API calls, model usage, token consumption, and costs. Data is aggregated at the hourly level - we recommend polling this endpoint at most once per hour. Rate limited to 20 requests per minute per team.
Cursor AI Code Changes — Retrieve granular accepted AI changes, grouped by deterministic changeId. Useful to analyze accepted AI events independent of commits. Returns changeId, user info, source (TAB/COMPOSER), model, line counts, and file metadata.
Cursor AI Code Changes CSV — Download change metrics data in CSV format for large data extractions. Streams in pages of 10,000 records server-side. Returns CSV data with columns: change_id, user_id, user_email, source, model, total_lines_added, total_lines_deleted, created_at, metadata_json.
Cursor AI Code Commits — Retrieve aggregated per-commit metrics that attribute lines to TAB, COMPOSER, and non-AI. Returns commit hash, user info, repository/branch info, line counts by source (TAB/COMPOSER/non-AI), commit message, and timestamps.
Cursor AI Code Commits CSV — Download commit metrics data in CSV format for large data extractions. Streams in pages of 10,000 records server-side. Returns CSV data with columns: commit_hash, user_id, user_email, repo_name, branch_name, is_primary_branch, total_lines_added, total_lines_deleted, tab_lines_added, tab_lines_deleted, composer_lines_added, composer_lines_deleted, non_ai_lines_added, non_ai_lines_deleted, message, commit_ts, created_at.
Cursor Analytics Agent Edits — Get metrics on AI-suggested code edits accepted by your team in the Cursor IDE. Returns daily metrics including suggested/accepted/rejected diffs, lines added/deleted, and acceptance ratios.
Cursor Analytics By User Agent Edits — Get agent edits metrics organized by individual users with pagination support. Returns data keyed by user email addresses, each containing an array of that user's metrics.
Cursor Analytics By User Client Versions — Get client versions metrics organized by individual users with pagination support. Returns data keyed by user email addresses, each containing an array of that user's metrics.
Cursor Analytics By User Models — Get model usage metrics organized by individual users with pagination support. Returns data keyed by user email addresses, each containing an array of that user's metrics.
Cursor Analytics By User Tabs — Get Tab usage metrics organized by individual users with pagination support. Returns data keyed by user email addresses, each containing an array of that user's metrics.
Cursor Analytics By User Top File Extensions — Get top file extensions metrics organized by individual users with pagination support. Returns data keyed by user email addresses, each containing an array of that user's metrics.
Cursor Analytics Client Versions — Get distribution of Cursor IDE client versions used by your team (defaults to last 7 days). Reports the latest version for each user per day.
Cursor Analytics Dau — Get daily active user counts for your team. DAU is the number of unique users who have used Cursor in a given day. Returns DAU breakdown metrics for the Cursor CLI, Cloud Agents, and BugBot.
Cursor Analytics Leaderboard — Get a leaderboard of team members ranked by AI usage metrics. Returns separate leaderboards for Tab autocomplete and Agent edits. Without user filtering: Returns top 10 users. With user filtering: Returns top 10 PLUS specified users (with their actual rankings).
Cursor Analytics Models — Get metrics on AI model usage across your team. Returns daily breakdown of model usage including message counts and user counts per model.
Cursor Analytics Tabs — Get metrics on Tab autocomplete usage across your team. Returns daily metrics including suggestions, accepts, rejects, and lines added/deleted via Tab completions.
Showing the first 25 tools — full list and schemas: Directory API.
How many tools are included?
The bundle exposes the full Cursor API surface on MCPBundles—Cloud Agents lifecycle, generic fetch/delete helpers, team admin, analytics, and AI-code tracking—within one connected server.
How do I connect Cursor to my AI agent?
Add the MCPBundles server URL to your MCP client configuration (Claude Desktop, Cursor, VS Code, etc.). The URL format is: https://mcp.mcpbundles.com/bundle/cursor. Authentication is handled automatically.
How many tools does Cursor provide?
Cursor provides 33 tools that can be called by AI agents, along with a SKILL.md that gives your AI agent domain knowledge about when and how to use them.
What authentication does Cursor require?
Cursor uses API Key. Cursor requires credentials. Connect via MCPBundles and authentication is handled automatically.
MCPBundles is an independent platform built on the open Model Context Protocol standard. MCPBundles is not affiliated with, sponsored by, or endorsed by Anthropic PBC.
More in developer tools
Other MCP servers in this category from the directory index