Main Menu
  • Tools
  • Developers
  • Topics
  • Discussions
  • News
  • Blogs
  • Builds
  • Contests
  • Compare
  • Arena
Create
    EveryDev.ai
    Sign inSubscribe
    Home
    Tools

    2,038+ AI tools

    • New
    • Trending
    • Featured
    • Compare
    • Arena
    Categories
    • Agents1104
    • Coding995
    • Infrastructure429
    • Marketing408
    • Design354
    • Projects323
    • Analytics311
    • Research297
    • Testing194
    • Data166
    • Integration164
    • Security162
    • MCP152
    • Learning143
    • Communication126
    • Extensions118
    • Commerce112
    • Prompts109
    • Voice105
    • DevOps89
    • Web73
    • Finance19
    1. Home
    2. Tools
    3. CrabTrap
    CrabTrap icon

    CrabTrap

    Application Security
    Featured

    An HTTP/HTTPS proxy that sits between AI agents and external APIs, evaluating every outbound request against security policies before it reaches the internet.

    Visit Website

    At a Glance

    Pricing
    Open Source

    Fully free and open-source under the MIT License. Self-host via Docker alongside PostgreSQL.

    Engagement

    Available On

    Linux
    Web
    API
    CLI

    Resources

    WebsiteDocsGitHubllms.txt

    Topics

    Application SecurityAutonomous SystemsNetwork Security

    Alternatives

    Anthropic Cybersecurity SkillsOuttakeCodeWall
    Developer
    BrexSan Francisco, CAEst. 2017$1.5B+ raised

    Listed Apr 2026

    About CrabTrap

    CrabTrap is an open-source HTTP/HTTPS forward proxy designed to secure AI agents that call external services like Slack, Gmail, and GitHub. It intercepts every outbound request, evaluates it against deterministic static rules and an LLM-based policy judge, and either forwards or blocks it with a reason. Every request and decision is logged to PostgreSQL for a complete audit trail. Built by Brex, it runs as a Docker container and provides a web UI for policy management and audit review.

    • HTTPS Interception — transparent MITM proxy with custom TLS server certificate generation per host, enabling full request inspection
    • Two-Tier Policy Evaluation — deterministic static rules (prefix, exact, glob URL patterns with optional HTTP method filters) are checked first; the LLM judge is only invoked if no rule matches
    • SSRF Protection — blocks requests to private networks (RFC 1918, loopback, link-local, CGN, IPv6 ULA/NAT64/6to4) with DNS-rebinding prevention
    • Per-Agent LLM Policies — natural-language security policies evaluated via LLM judge with configurable fallback (deny or passthrough) when the judge is unavailable
    • Circuit Breaker — trips after 5 consecutive LLM failures and reopens after a 10-second cooldown to maintain proxy availability
    • Per-IP Rate Limiting — token bucket rate limiter with a default of 50 req/s and burst of 100
    • Prompt Injection Defense — request payloads are JSON-encoded and policy content is JSON-escaped before being sent to the LLM judge
    • Policy Builder — an agentic loop that analyzes observed traffic and automatically drafts security policies
    • Eval System — replay historical audit log entries against a policy to measure accuracy before deploying
    • Web UI — audit trail viewer, policy editor, eval results, and agent management accessible at localhost:8081
    • PostgreSQL Audit Trail — every request, decision, and response is recorded for compliance and forensic review
    • Docker-Based Deployment — runs via docker compose up -d alongside PostgreSQL; cross-platform binaries available for linux/darwin on amd64/arm64
    CrabTrap - 1

    Community Discussions

    Be the first to start a conversation about CrabTrap

    Share your experience with CrabTrap, ask questions, or help others learn from your insights.

    Pricing

    OPEN SOURCE

    Open Source (MIT)

    Fully free and open-source under the MIT License. Self-host via Docker alongside PostgreSQL.

    • HTTPS interception and MITM proxy
    • Static rule engine (prefix, exact, glob)
    • LLM-based policy judge
    • SSRF protection
    • Per-IP rate limiting

    Capabilities

    Key Features

    • HTTPS interception via transparent MITM proxy
    • Two-tier policy evaluation (static rules + LLM judge)
    • SSRF protection with DNS-rebinding prevention
    • Per-agent natural-language LLM security policies
    • Circuit breaker for LLM judge failures
    • Per-IP token bucket rate limiting
    • Prompt injection defense
    • Agentic policy builder from observed traffic
    • Eval system for replaying audit entries against policies
    • Web UI for audit trail, policy editor, and agent management
    • PostgreSQL audit trail for all requests and decisions
    • Docker-based deployment with multi-arch images

    Integrations

    PostgreSQL
    Docker
    Slack
    Gmail
    GitHub
    GoReleaser
    React
    TypeScript
    Vite
    API Available
    View Docs

    Reviews & Ratings

    No ratings yet

    Be the first to rate CrabTrap and help others make informed decisions.

    Developer

    Brex

    Brex builds financial services and software for innovative companies, integrating corporate cards, expense management, and banking into a unified platform. Founded by entrepreneurs and backed by investors including Y Combinator, Greenoaks, and Tiger Global, the company serves over 35,000 businesses from startups to enterprises. Brex empowers employees to make better financial decisions while helping companies launch confidently and scale smarter.

    Founded 2017
    San Francisco, CA
    $1.5B+ raised
    1,200 employees

    Used by

    OpenAI
    Airbnb
    DoorDash
    Coinbase
    +4 more
    Read more about Brex
    WebsiteLinkedIn
    2 tools in directory

    Similar Tools

    Anthropic Cybersecurity Skills icon

    Anthropic Cybersecurity Skills

    An open-source library of 754 structured cybersecurity skills for AI agents, mapped to 5 frameworks across 26 security domains.

    Outtake icon

    Outtake

    AI-powered digital risk protection platform that detects and dismantles impersonation attacks across domains, social media, apps, and advertisements.

    CodeWall icon

    CodeWall

    AI-powered autonomous pentesting platform that continuously attacks your infrastructure, chains real exploits, and delivers verified remediation.

    Browse all tools

    Related Topics

    Application Security

    AI tools for securing software applications and identifying vulnerabilities.

    54 tools

    Autonomous Systems

    AI agents that can perform complex tasks with minimal human guidance.

    149 tools

    Network Security

    Tools for securing networks, including VPNs and firewalls.

    7 tools
    Browse all topics
    Back to all tools
    Explore AI Tools
    • AI Coding Assistants
    • Agent Frameworks
    • MCP Servers
    • AI Prompt Tools
    • Vibe Coding Tools
    • AI Design Tools
    • AI Database Tools
    • AI Website Builders
    • AI Testing Tools
    • LLM Evaluations
    Follow Us
    • X / Twitter
    • LinkedIn
    • Reddit
    • Discord
    • Threads
    • Bluesky
    • Mastodon
    • YouTube
    • GitHub
    • Instagram
    Get Started
    • About
    • Editorial Standards
    • Corrections & Disclosures
    • Community Guidelines
    • Advertise
    • Contact Us
    • Newsletter
    • Submit a Tool
    • Start a Discussion
    • Write A Blog
    • Share A Build
    • Terms of Service
    • Privacy Policy
    Explore with AI
    • ChatGPT
    • Gemini
    • Claude
    • Grok
    • Perplexity
    Agent Experience
    • llms.txt
    Theme
    With AI, Everyone is a Dev. EveryDev.ai © 2026
    Discussions