BugBug MCP server
Commit Nov 12, 2026Enables AI agents to automate test creation, debug failed runs, and manage test suites using live BugBug project data and workflow guidance.
Also known as bugbug-io/agent-plugin, agent-plugin
- Stars
- 4
- Tools
- —
- Downloads, 30 days
- —
- Latest version
- 14.23.1
- Last commit
- Nov 12, 2026
- License
- MIT
What is BugBug MCP?
BugBug enables AI agents to automate test creation, debug failed runs, and manage test suites using live BugBug project data and workflow guidance.
Summary generated from the project's README
About BugBug MCP server
BugBug provides an MCP connection that allows AI agents to interact directly with live project data for automated testing. Developers can integrate this server into clients like Cursor, Claude, VS Code, Codex, or GitHub Copilot using a single command via the BugBug CLI. Once installed, the selected AI client handles the authentication through an OAuth sign-in process to connect to the production endpoint.
The server combines 50 MCP tools with 9 specialized skills and 8 workflow prompts to assist with the entire testing lifecycle. It enables agents to plan coverage, author YAML, and manage element selectors. While skills provide workflow guidance, the MCP connection supplies real-time data such as run history, failure diagnostics, screenshots, and DOM snapshots. Users on Pro plans or higher can access the MCP features, while skills remain available on all plans. Any changes to BugBug resources require explicit user authorization.
What you can do with it
- 1Automate test coverage planningAsk your AI agent to create a plan for end-to-end and UI test coverage based on project requirements.
- 2Debug failed test runsInstruct the agent to investigate why a specific test run failed using available diagnostics and screenshots.
- 3Maintain and repair selectorsHave the AI review and optimize element selectors to ensure test reliability.
- 4Generate project health reportsRequest a read-only summary of project health to provide insights for product managers and QA teams.
Written from the project's README. Every use case is checked against a quote from it.
How to install BugBug MCP
Formats verified 10 Nov 2026Install BugBug MCP in Claude Code
$ claude mcp add --transport http agent-plugin https://mcp.bugbug.io/mcpInstall BugBug MCP in Cursor
~/.cursor/mcp.json
{
"mcpServers": {
"agent-plugin": {
"url": "https://mcp.bugbug.io/mcp"
}
}
}Install BugBug MCP in VS Code
.vscode/mcp.json
{
"servers": {
"agent-plugin": {
"type": "http",
"url": "https://mcp.bugbug.io/mcp"
}
}
}Install BugBug MCP in Codex
~/.codex/config.toml
[mcp_servers.agent-plugin] url = "https://mcp.bugbug.io/mcp"
Install BugBug MCP in Gemini CLI
~/.gemini/settings.json
{
"mcpServers": {
"agent-plugin": {
"httpUrl": "https://mcp.bugbug.io/mcp"
}
}
}Automated checks
Ran Nov 14, 2026Full report →Not verified yet. A server needs a score of 80, no serious finding, and enough data for us to check. How checks work
- Source repository linked and reachablebugbug-io/agent-plugin is public and reachablegithub.com/bugbug-io/agent-plugin ↗Pass
- Package names this repository as its sourceNot checked
- Build provenance (Sigstore attestation)Not checked
- Open source license (OSI)MITspdx.org/licenses/MIT.html ↗Pass
- Commit or release in the last yearLast commit or release 2026-11-12github.com/bugbug-io/agent-plugin ↗Pass
- OpenSSF ScorecardNot checked
- Package name not close to a popular packageNot checked
- No critical advisory in the package or its dependenciesNot checked
- No hidden instructions to the model0 tool descriptions and the README read; no override phrases, credential paths or hidden charactersPass
- Destructive tools are labelledNot checked
- Remote endpoint answers the MCP handshakeHTTPS endpoint answered 401 (sign-in required)Pass
README
View on GitHub
BugBug Agent Plugin
Install BugBug's skills and hosted MCP connection into your AI client with the BugBug CLI:
npx @bugbug-io/cli plugin --agent=<agent>
Choose exactly one supported client: cursor, claude, vscode, codex, or
copilot. Use --dry-run to preview the installation.
The installer does not accept --token. The AI client you select completes its
own MCP OAuth sign-in after installation. The plugin connects to the production
endpoint at https://mcp.bugbug.io/mcp.
Features
- Ready in one command — Set up Claude Code, Codex, Cursor, GitHub Copilot, or VS Code with the BugBug CLI; the selected client completes OAuth sign-in.
- Turn requirements into coverage — Use BugBug's workflow guidance and live project context to plan coverage, then create tests from an approved plan.
- Build with live BugBug context — Combine 50 MCP tools and 8 workflow prompts with 9 skills so your agent can inspect the tests, suites, profiles, and project details it needs.
- Debug from evidence — Investigate failed runs with run details, failure diagnostics, screenshots, and DOM snapshots, guided by a safe debugging workflow.
- Keep suites reliable as they grow — Review selectors, tests, reusable components, and suites with BugBug-specific guidance before making targeted improvements.
- Turn run history into decisions — Summarize project and run health for actionable QA and product insights, using live data instead of guesswork.
Skills provide safe workflow guidance; MCP supplies the live BugBug data and actions. Changes to BugBug resources require the user's explicit authorization.
What you get
Skills are available on all BugBug plans. MCP access requires Pro or higher.
Bundled skills
| Skill | Capability |
|---|---|
bugbug-context-discovering |
Discover the BugBug project, repository, and tested-application context. |
bugbug-project-reporting |
Create read-only project health reports for PM and QA audiences. |
bugbug-selectors-authoring |
Author, review, repair, and optimize element selectors. |
bugbug-steps-authoring |
Choose step types and create or update step payloads. |
bugbug-test-run-debugging |
Investigate test runs and classify failures. |
bugbug-tests-authoring |
Create and maintain tests from an approved plan. |
bugbug-tests-planning |
Plan end-to-end and UI test coverage. |
bugbug-tests-refactoring |
Analyze and refactor tests, groups, components, and suites. |
bugbug-yaml-authoring |
Create, validate, repair, and explain BugBug YAML. |
When to use it
Use the plugin when you want an AI coding client to work with BugBug tests and test results. Use its skills for workflow guidance and its MCP connection for live BugBug data, runs, exports, diagnostics, and explicit mutations.
For the complete MCP registry, see the MCP server docs.
BugBug MCP questions
How do I add BugBug MCP to Claude Code?
Run claude mcp add --transport http agent-plugin https://mcp.bugbug.io/mcp in your terminal. Add --scope user to make it available in every project.
How do I add BugBug MCP to Cursor?
Open ~/.cursor/mcp.json and add the server under mcpServers, using the Cursor config shown in the install section above.
Which AI clients are supported?
The plugin supports Claude Code, Codex, Cursor, GitHub Copilot, and VS Code.
How do I authenticate with the BugBug MCP server?
The selected AI client handles the MCP OAuth sign-in process automatically after installation.
Does the MCP connection require a specific plan?
While skills are available on all plans, accessing the MCP connection requires a Pro plan or higher.
Can the AI agent make changes to my BugBug resources?
Yes, but any changes to BugBug resources require explicit authorization from the user.