The Burp Suite for AI Agent MCP server bridges Burp Suite and AI agents, enabling automated penetration testing by capturing and analyzing HTTP traffic, managing tasks and findings, and interacting with Burp tools. Key capabilities include:
Monitoring bridge connection and statistics (burp_status)
Listing and filtering captured HTTP requests (burp_requests)
Viewing full request/response details (burp_request_detail)
Enumerating unique endpoints with parameters and hit counts (burp_endpoints)
Listing scan/plan/scope tasks from Burp (burp_tasks)
Viewing and importing security findings as Burp issues (burp_issues, burp_import_issue)
Retrieving session snapshots (cookies, storage) (burp_snapshot)
Sending actions to Burp: open in Repeater, add scan issue, log messages (burp_send_to_burp)
Checking pending outbound actions (burp_outbound_status)
Clearing all captured data (burp_clear)
Provides tools to capture HTTP requests, analyze endpoints, queue security scans, import issues, and send actions to Burp Suite (Repeater, Scanner, etc.) through a two-way MCP bridge.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Burp Suite for AI Agent list captured requests"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
burpsuite-for-ai-agent/
āāā src/
ā āāā index.js ā MCP server (HTTP bridge + 11 MCP tools)
āāā plugin/
ā āāā burpAI.py ā Jython 2.7 Burp plugin (context menu, auto-forward, outbound polling)
āāā package.json ā Node.js manifest (dependencies: @modelcontextprotocol/sdk, zod)
āāā .gitignore
āāā README.mdRelated MCP server: novelty
Burp Suite (Linux / Windows)
āā plugin/burpAI.py (Jython plugin)
ā POST /ingest (127.0.0.1:9999)
ā GET /burp/outbound (poll every 2s)
ā¼
src/index.js ā MCP server (HTTP bridge + 11 MCP tools)
ā
ā¼
AI Agent āā MCP toolsLinux: Burp + server + agent ā all on one machine. Windows + WSL: Burp on Windows, server + agent in WSL over localhost.
bash install.shThis runs npm install, prints Burp plugin instructions, and generates .mcp.json for AI agent auto-registration.
node src/index.jsBurp Suite ā Extensions ā Installed ā Add
Extension type: Python (requires Jython 2.7 standalone JAR)
File: plugin/burpAI.py
In Burp's burpAI tab, click Check Status. Connected:
{ "ok": true, "requests": 0, "endpoints": 0 }Tool | Description |
| Bridge connection status + store statistics |
| List captured HTTP requests (filter by url/method) |
| Full request/response (headers, body) |
| Unique endpoints with parameter names, hit counts |
| Scan/plan/scope tasks queued from Burp context menu |
| Security findings queued for Burp import |
| Submit a finding (title, url, severity, detail) |
| Latest session snapshot (cookies, storage) |
| Queue action: send_to_repeater, add_scan_issue, console_log |
| Modify captured request ā send to Burp Repeater |
| Pending outbound actions |
| Run nuclei scan against a single URL |
| Run nuclei scan against all captured endpoints |
| List/display nuclei scan results |
| Import nuclei findings as Burp issues |
| Clear all captured data |
Right-click a request in Burp (Proxy/Repeater) ā burpAI: send request(s)
Agent:
burp_requests ā list captured requests
burp_endpoints ā enumerate endpoints
burp_request_detail ā full request/response
Right-click ā burpAI: send + queue scan
Agent picks up via burp_tasks.
Agent calls burp_import_issue ā Burp tab ā Import Issues
Agent calls burp_send_to_burp with type send_to_repeater ā Burp opens Repeater tab.
This is the core custom scanner flow: Burp captures ā agent thinks ā agent modifies ā Burp executes.
Burp (Proxy ā History)
ā right-click ā burpAI: send request(s)
ā¼
MCP server āāā captures request
ā
ā¼
Agent reads via burp_requests / burp_request_detail
ā
ā Agent thinks: "endpoint /api/login takes username, try SQLi"
ā
ā¼
Agent calls burp_replay({
request_id: "burp:burp-12345",
set_body: '{"username":"admin\\' OR 1=1--","password":"x"}'
})
ā
ā¼
MCP server ā Burp Repeater ā Burp executes HTTP call
ā
ā¼ (auto-forward enabled)
Response captured back ā Agent reads & analyzes
ā
ā¼
burp_import_issue ā if vulnerableKey tool: burp_replay takes a captured request ID, applies modifications (method, path, headers, body), and sends it to Burp Repeater. No need to manually craft raw bytes ā the agent just specifies what to change.
Requirements:
Enable Auto-send Repeater responses in Burp plugin settings
This ensures the agent sees every Repeater response automatically
Optionally run nuclei scans directly from the bridge:
Tool | Use case |
| Scan one URL with nuclei templates |
| Scan all captured endpoints at once |
| Review scan history |
| Import nuclei findings as Burp issues |
Nuclei must be installed on the same machine. Findings with High/Critical severity are auto-imported as Burp issues.
Feature | Description |
Auto-send Proxy responses | All Proxy traffic sent to MCP server |
Auto-send Repeater responses | Repeater traffic automatically forwarded |
Forward Burp Scanner issues | Scanner findings pushed to agent |
Auto import issues | Pull agent findings on context menu |
node src/index.js --port 9999 # Custom port (default: 9999)
node src/index.js --db /path/to/data.json # Persistence file (default: burpai-data.json)
node src/index.js --port 9001 --db custom.jsonData persists across restarts when --db is set (or by default). The server auto-saves on every mutation and loads data on startup.
Burp plugin URL configurable from the burpAI settings tab in Burp.
Node.js 18+
Burp Suite (Community or Professional)
Jython 2.7 standalone JAR (configured in Burp Extensions ā Environment)
Clear all captured requests, endpoints, tasks, and issues from the bridge store.
Inputs: no named parameters.
List unique endpoints (METHOD + path) observed from Burp traffic, with query/body parameter names and hit counts. Helps identify attack surface.
Inputs: method, url_contains.
Submit a confirmed finding as a Burp-importable issue. The issue will appear in the bridge and can be pulled into Burp via the burpAI Burp plugin.
Inputs: url (required), title (required), detail (required), method, +5 more.
List security findings/issues queued for Burp import. These are confirmed findings that can be imported as Burp Scanner issues.
Inputs: no named parameters.
List pending outbound actions queued for the Burp plugin to execute. Actions remain here until the Burp plugin polls and drains them.
Inputs: no named parameters.
Fetch full details for one captured request: headers, request body, and response body (when available). Pass the id from burp_requests.
Inputs: id (required), body_max_chars.
List recent HTTP requests captured from Burp (most-recent first). Returns id, method, url, and status for each captured request.
Inputs: limit, method, url_contains.
Queue an action for execution in Burp Suite. The next time the Burp plugin polls, it will execute the action. Supports: send_to_repeater (open Repeater tab with a request), add_scan_issue (add scan issue to Burp), console_log (write to Burp output tab).
Inputs: type (required), url, host, port, +7 more.
Return the most recent session snapshot (cookies, localStorage, sessionStorage). Useful for constructing authenticated requests.
Inputs: url_contains.
Show Burp bridge connection status and store statistics (requests, tasks, issues captured). Call first to confirm bridge is running.
Inputs: no named parameters.
List scan / plan / scope tasks queued from the Burp extension. Use these to decide what to analyze next.
Inputs: no named parameters.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelogā¦
Scored across 11 tools
Each tool targets a distinct aspect of the Burp Suite integration: clearing, status, requests (list/detail), endpoints, tasks, issues (list/import), session snapshot, and outbound actions (send/status). Despite some related operations, the descriptions clearly differentiate them, leaving no ambiguity.
All tools follow a consistent 'burp_' prefix with snake_case names. Verbs and nouns are used predictably (e.g., burp_clear, burp_requests, burp_import_issue). There is no mixing of conventions, ensuring easy pattern recognition.
With 11 tools, the set is well-scoped for a Burp Suite bridge. It covers essential operations without being bloated or sparse. Each tool serves a clear purpose, and the count aligns with the expected complexity of security testing workflows.
The tool set covers the core workflow: connection status, request capture, endpoint discovery, task and issue management, import, session retrieval, and outbound actions. Minor gaps exist, such as the lack of detailed views for individual tasks or issues, but these do not critically hinder the primary use case.
Zero-setup MCP gateway securely connecting AI to your tools with authentication and workflows
MCP server for Pentest-Tools.com: run scans, manage findings and reports via your preffered LLM.
Talk to your public-facing AI from any MCP client ā Claude, ChatGPT, Cursor, Cline, Windsurf.
Security scanner for AI agents and MCP servers: injection, jailbreak, tool abuse.
