Atlas / MCP servers / fuzzinglabs / Security Hub

Security HubSAFE

Name: Security Hub
Author: fuzzinglabs

mcp/fuzzinglabs/security-hub

MCP server by fuzzinglabs · ai · claude · cybersecurity · Audited 2026-09-27

A growing collection of MCP servers bringing offensive security tools to AI assistants. Nmap, Ghidra, Nuclei, SQLMap, Hashcat and more.

Verdict

SAFE

Grade

Trust score

89 /100

Exposed tools

69 61r · 8w · 0d

Transport

stdio

License

MIT

Stars

794

Overview

From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.

[](https://github.com/FuzzingLabs/mcp-security-hub/actions/workflows/build.yml) [](https://github.com/FuzzingLabs/mcp-security-hub/actions/workflows/security-scan.yml) [](https://opensource.org/licenses/MIT) [](https://modelcontextprotocol.io/)

Production-ready, Dockerized MCP (Model Context Protocol) servers for offensive security tools. Enable AI assistants like Claude to perform security assessments, vulnerability scanning, and binary analysis.

Features

38 MCP Servers covering reconnaissance, web security, binary analysis, blockchain security, cloud security, code security, secrets detection, threat intelligence, OSINT, Active Directory, fuzzing, and more
300+ Security Tools accessible via natural language through Claude or other MCP clients
Production Hardened - Non-root containers, minimal images, Trivy-scanned
Docker Compose orchestration for multi-tool workflows
CI/CD Ready with GitHub Actions for automated builds and security scanning

Quick Start

# Clone the repository
git clone https://github.com/FuzzingLabs/mcp-security-hub
cd mcp-security-hub

# Build all MCP servers
docker-compose build

# Start specific servers
docker-compose up nmap-mcp nuclei-mcp -d

# Verify health
docker-compose ps

Configure Claude Desktop / Claude Code

Important: You must build the image

Read from source at commit dddbfe993d4dOBSERVED · 2026-09-27

Exposed tools (69)

61 read · 8 write · 0 destructive.

Tool	Risk	Description
`binwalk_entropy`	read	Analyze file entropy to detect compressed or encrypted sections.
`binwalk_extract`	read	Extract embedded files and filesystems from firmware.
`binwalk_hexdump`	read	Display hex dump of a file section. Useful for examining
`binwalk_scan`	read	Scan a firmware file for embedded files, filesystems, and signatures.
`boofuzz_create_script`	write	Create and save a Boofuzz Python script on the server. The script should accept
`boofuzz_get_results`	read	Retrieve the crash log or audit results from a fuzzing session.
`boofuzz_list_scripts`	read	List all saved fuzzer scripts.
`boofuzz_run_fuzzer`	write	Execute a saved Boofuzz fuzzer script against a target. Runs asynchronously with a timeout.
`capa_analyze`	read	Analyze a binary file for capabilities using Mandiant
`damlviewer_generate_table`	read	Generate a DAML access-control markdown table from a project directory or a single `.daml` file.
`dharma_generate`	read	Generate test cases using a Dharma grammar file located on the server.
`dharma_generate_custom`	read	Generate test cases from a custom Dharma grammar provided as a string. Useful for dynamically generated grammars or quick testing without saving files.
`fetch_wayback_urls`	read	Fetch all URLs from the Wayback Machine for a given domain.
`ffuf_custom`	read	Custom fuzzing with full control over all options.
`ffuf_dir`	read	Directory and file discovery fuzzing.
`ffuf_param`	read	Parameter fuzzing.
`ffuf_vhost`	read	Virtual host discovery.
`get_analysis_results`	read	Retrieve results from a previous analysis.
`get_fetch_results`	read	Retrieve results from a previous waybackurls fetch by fetch ID.
`get_fuzz_results`	read	Retrieve results from a previous fuzzing session by scan ID.
`get_run_results`	write	Retrieve results from a previous run by run ID.
`get_scan_results`	read	Retrieve results from a previous scan by scan ID.
`gitleaks_detect`	read	Quick scan provided content (text/code) for secrets.
`gitleaks_scan_dir`	read	Scan a directory for secrets without git history analysis.
`gitleaks_scan_repo`	read	Scan a git repository for secrets and credentials.
`list_active_fetches`	read	List currently running waybackurls fetches.
`list_active_runs`	read	List currently running DAML viewer jobs.
`list_active_scans`	read	List currently running scans.
`list_checks`	read	List available security checks for a cloud provider.
`list_compliance_frameworks`	read	List available compliance frameworks for each cloud provider.
`list_extractions`	read	List all completed extractions with their file contents.
`list_recent_searches`	read	List recent exploit searches.
`list_rulesets`	read	List available YARA rule sets and their details.
`list_runs`	read	List completed DAML viewer runs (most recent first).
`list_templates`	read	List available Nuclei template categories and common tags.
`list_wordlists`	read	List available wordlists for fuzzing.
`masscan_scan`	read	Fast port scan using masscan. Can scan millions of hosts per minute.
`masscan_top_ports`	read	Scan the top 20 most common ports on target(s).
`medusa_fuzz`	write	Run the Medusa fuzzing process.
`medusa_get_config`	read	Read medusa.json.
`medusa_init`	read	Initialize Medusa in the current directory.
`medusa_update_config`	write	Update medusa.json fields.
`nuclei_scan`	read	Comprehensive vulnerability scan using Nuclei templates.
`os_detection`	read	Fingerprint the operating system of the target.
`port_scan`	read	Scan ports on a target host or network.
`prowler_compliance`	read	Check cloud environment against compliance frameworks
`prowler_scan`	write	Run security assessment on AWS, Azure, GCP, or Kubernetes.
`quick_scan`	read	Fast scan of the 100 most common ports with service detection.
`script_scan`	write	Run NSE (Nmap Scripting Engine) scripts against target.
`searchsploit_examine`	read	Get the code/content of a specific exploit by EDB-ID.
`searchsploit_search`	read	Search Exploit-DB for exploits by keyword.
`service_scan`	read	Detect service versions on target ports.
`solazy_dotting`	read	Reinsert functions (clusters) into a reduced CFG `.dot` using a JSON config.
`solazy_fetcher`	read	Fetch a deployed program bytecode (or account) from a Solana RPC endpoint.
`solazy_recap`	read	Generate an audit-friendly markdown recap for an Anchor project.
`solazy_reverse`	read	Reverse engineer a compiled Solana sBPF `.so` file.
`solazy_sast`	write	Run SAST (Starlark rules) on an Anchor or SBF project directory.
`sql_dump`	read	Dump data from database tables. Use responsibly and only on authorized targets.
`sql_enumerate`	read	Enumerate databases, tables, or columns from a vulnerable target.
`sql_scan`	read	Scan a URL for SQL injection vulnerabilities.
`sql_test`	read	Quick test to check if a parameter is injectable.
`template_scan`	read	Scan with specific template categories.
`trivy_generate_sbom`	read	Generate a Software Bill of Materials (SBOM) for an image.
`trivy_scan_config`	read	Scan IaC files (Terraform, Dockerfile, Kubernetes YAML, etc.)
`trivy_scan_filesystem`	read	Scan a filesystem path or git repository for vulnerabilities
`trivy_scan_image`	read	Scan a container image for vulnerabilities.
`whatweb_scan`	read	Identify web technologies, CMS, frameworks, and server software.
`yara_scan`	read	Scan files or directories using YARA rules for malware detection
`yara_scan_with_rules`	read	Scan files using custom inline YARA rules.

Trust audit

SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.

Layer	What it checks	Result
L0	Provenance & inventory	PASS
L1	Static analysis of the code	WARN
L2	Instruction surface (what it tells the agent)	PASS
L3	Class-specific surface	PASS
L4	Behavioural (sandbox)	SKIPPED

What the source does

Filesystem: declared (5 observation(s))
Network: declared (6 observation(s))
Shell: none-observed
Dependencies: not all pinned
Secrets in source: none-found

Findings (8)

MEDIUMNetwork egress · net.raw_ip · CWE-200, CWE-319

web-security/ffuf-mcp/server.py:362

"description": "Target URL (e.g., https://10.10.10.10)",

LOWInventory / provenance · inv.hidden_file · CWE-1104

.trivyignore

.trivyignore

Why it matters. hidden member outside the usual dotfiles

Fix. review its purpose

LOWNetwork egress · net.raw_ip · CWE-200, CWE-319

web-security/burp-mcp/README.md:79

2. Configure Claude Desktop to connect to `http://127.0.0.1:9876`

LOWSupply chain · supply.unpinned · CWE-829, CWE-1357

binary-analysis/binwalk-mcp/requirements.txt

mcp, pydantic, pydantic-settings