MCP Servers · six2dez

burp-ai-agent

Custom AI Agent is a Burp Suite extension that integrates multiple AI backends (local and cloud) to assist with vulnerability analysis and security testing. It provides passive/active scanning with AI-assisted insights, MCP tool integration for external AI agents, and privacy controls for handling sensitive data during analysis.

Talk to DEV.co Discuss Implementation

Source: GitHub — github.com/six2dez/burp-ai-agent

1.3k

GitHub stars

203

Forks

Kotlin

Primary language

MIT

License (OSI-approved)

Key facts

Objective fields from the source. Values we can't verify are shown as “Unknown” rather than guessed.

Field	Value
Repository	six2dez/burp-ai-agent
Owner	six2dez
Primary language	Kotlin
License	MIT — OSI-approved
Stars	1.3k
Forks	203
Open issues	9
Latest release	v0.9.0 (2026-06-26)
Last updated	2026-07-01
Source	https://github.com/six2dez/burp-ai-agent

What burp-ai-agent is

A Kotlin-based Burp extension that bridges Burp Suite with AI via 11 backends (Ollama, OpenAI-compatible, Anthropic, Claude CLI, etc.), exposes 59 MCP tools for autonomous scanning, implements AES-256-GCM secrets encryption, HKDF host anonymization, ReDoS-safe redaction, and token-budget guardrails. Supports passive scanner integration (Burp Pro), external MCP servers (SSE/stdio), and audit logging (JSONL with SHA-256).

Quickstart

Get the burp-ai-agent source

Clone the repository and explore it locally.

terminalbash

1git clone https://github.com/six2dez/burp-ai-agent.git2cd burp-ai-agent3# follow the project's README for install & configuration

Need it deployed, integrated, or customized instead? DEV.co ships production installs.

Best use cases

AI-Augmented Penetration Testing

Analysts can run passive/active scans with AI reasoning over Burp traffic in real time, then manually refine findings or launch autonomous scanning via Claude Desktop MCP.

Security Team Workflow Automation

Teams can connect their own AI models (Ollama, LM Studio, or cloud APIs) to Burp, build custom prompt libraries, and audit all AI interactions with structured JSONL logs.

AppSec Integration & CI/CD

Organizations can use the Burp Scan Skill with terminal-based AI assistants (Claude Code, Gemini CLI) to programmatically scan endpoints, correlate 62 vulnerability classes, and generate compliance-ready audit trails.

Implementation considerations

Requires Java 21; Burp Suite Community or Pro. Install via BApp Store or build from source with Gradle shadowJar (2 variants: full 59 MCP tools or store build 8 tools).
AI backend selection is critical: local models (Ollama/LM Studio) have zero cloud overhead but require pre-downloaded models and local inference hardware; cloud APIs (Anthropic, Perplexity) offer capability but add latency and per-token costs.
MCP server setup for Claude Desktop requires Node.js 18+ and supergateway CLI; external MCP servers add topology complexity (SSE/stdio transports, auth token storage, per-host scope enforcement).
Passive scanner runs in Burp Pro only; audit logging (JSONL + SHA-256 hashes) provides compliance trail but requires manual collection and analysis for multi-user teams.
Secrets encryption uses per-install keys stored locally; no cloud key management or key rotation policy documented. Review threat model for your threat level.

When to avoid it — and what to weigh

Burp Community Edition with Advanced Scanning — The passive scanner requires Burp Pro. Community users get only manual request analysis; no automatic vulnerability detection.
Air-Gapped or Strict Network Isolation — Most backends (Anthropic, Perplexity, OpenAI-compatible cloud) require internet access. Local-only options (Ollama, LM Studio) exist but require prior model downloads and CPU/GPU resources.
Zero Tolerance for Plaintext Secrets in Code or Config — While v0.9.0 encrypts secrets at rest using per-install AES-256-GCM keys, API keys must still be entered into the extension UI. No HSM or external secret-management integration mentioned.
Minimal Customization or Locked-Down Deployment — Extension relies on user-supplied AI models and external MCP servers; limited guidance for sandboxing or enforcing fine-grained policy per MCP tool in enterprise environments.

License & commercial use

MIT License. Permissive; allows commercial use, modification, and distribution with attribution. No copyleft obligations.

MIT license is permissive for commercial use. However, integration of third-party AI backends (Anthropic, OpenAI, Perplexity, NVIDIA, etc.) introduces separate commercial terms: users must comply with each provider's ToS and licensing. No warranty or indemnification from the extension itself.

DEV.co evaluation signals

Editorial assessment — not user reviews. Directional, with an explicit confidence level.

Signal	Assessment
Maintenance	Active
Documentation	Strong
License clarity	Clear
Deployment complexity	Moderate
DEV.co fit	Strong
Assessment confidence	High

Security considerations

v0.9.0 implements AES-256-GCM per-install secret encryption (not plaintext), HKDF host anonymization in STRICT mode, request/response body redaction with ReDoS-safe patterns, and pre-send secret tripwire warnings with audit logs. Threat model considerations: (1) per-install encryption keys are stored locally—no remote key derivation; (2) MCP external server auth tokens are encrypted at rest but require secure channel setup (SSE over TLS); (3) audit logs are JSONL with SHA-256 but rely on file system integrity—no immutable log transport; (4) no mention of Burp session hijacking, MCP client impersonation, or supply-chain vetting for external MCP servers. Redaction patterns are user-configurable but ReDoS protection scope is not detailed. No vulnerability disclosure policy or security audit history provided.

Alternatives to consider

Burp Suite's Built-in Burp AI (Pro only)

PortSwigger's native AI integration; no extension install, tighter Burp API surface. Limited to Burp's curated models/backends; less customization than Custom AI Agent.

ZAP (OWASP) + Custom Scripting

Free, open-source alternative scanner with Python/JavaScript scripting for AI integration. Steeper learning curve; no native MCP, fewer AI backends out-of-the-box.

Semgrep + IDE AI (GitHub Copilot, Claude)

CI/CD-native SAST + IDE-resident AI. Complements Burp but does not replace dynamic scanning and manual testing workflows; best for code review, not runtime vulnerability analysis.

Software development agency

Build on burp-ai-agent with DEV.co software developers

Install Custom AI Agent from the BApp Store, configure your preferred AI backend, and start AI-assisted penetration testing in Burp Suite today.

Talk to DEV.co

Related open-source tools

Surfaced by semantic similarity across the DEV.co open-source index.

deepseek-pp fast-agent agent-scan

Related on DEV.co

Explore the category and the services that help you build with it.

MCP Servers Web Application Security Testing API Security & Testing Application Development with AI Assistance Talk to DEV.co

burp-ai-agent FAQ

Do I need a paid AI API to use this extension?

No. You can use local models (Ollama, LM Studio) for zero per-token cost, or connect free CLI tools (Claude CLI, Gemini CLI). Cloud APIs (Anthropic, Perplexity, OpenAI-compatible) incur per-token charges. Extension supports all 11 backends concurrently.

Can I use Custom AI Agent without Burp Pro?

Yes, for manual request analysis. The passive scanner (automatic vulnerability detection) requires Burp Pro. Active scanning and MCP tools work in Community.

Is my API key stored in plaintext?

No. v0.9.0 encrypts all stored API keys with AES-256-GCM using a per-install key. However, keys must be entered into the extension UI at runtime; review your threat model for plaintext-in-memory exposure.

Can external AI agents (Claude Desktop) access my out-of-scope targets?

By default, MCP tools can be scoped to in-scope hosts in Burp. However, if the MCP client (e.g., Claude Desktop) has direct network access, it may bypass Burp's scope. Review MCP server configuration and client network policy.

Software development & web development with DEV.co

DEV.co helps companies turn open-source tools like burp-ai-agent into production software. Our software development services cover the full lifecycle — architecture, web development, integration, and maintenance — delivered by software developers and web developers who ship. Engage our software development agency to implement or customize it for your mcp servers stack.

Integrate AI into Your Security Testing

Install Custom AI Agent from the BApp Store, configure your preferred AI backend, and start AI-assisted penetration testing in Burp Suite today.

Talk to DEV.co Schedule a Consultation