EveryDev.ai
Subscribe
Home
Tools

5,923+ AI tools

  • Featured
  • Gems
  • Trending
  • New
  • Rate tools
  • Compare
  • Arena
Categories
  • Agents2992
  • Coding2260
  • Infrastructure1068
  • Projects824
  • MCP795
  • Research744
  • Analytics567
  • Marketing527
  • Design489
  • Testing476
  • Data442
  • Security417
  • Integration288
  • Communication286
  • Prompts284
  • Extensions239
  • Voice214
  • DevOps191
  • Commerce184
  • Web128
  • Learning123
  • Finance44
AI Tools by Topic
  • 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
  • Users
  • Rate Tools
  • 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 ©
    1. Home
    2. Tools
    3. Pipelock
    Pipelock icon

    Pipelock

    MCP Tools
    Featured

    Open-source agent firewall that scans mediated HTTP, MCP, A2A and WebSocket traffic for secret leaks, prompt injection and tool poisoning and emits signed action receipts.

    Visit Website

    At a Glance

    Pricing
    Open Source

    Free Apache-2.0 licensed core firewall; run on as many agents as you want.

    Engagement

    Available On

    Windows
    macOS
    Linux
    Web
    API

    Resources

    WebsiteDocsGitHubllms.txt

    Topics

    MCP ToolsAI InfrastructureNetwork Security

    Alternatives

    Eidolon AIEveLiveKit
    Developer
    PipeLabAtlanta, GAEst. 2024

    Listed Nov 2026

    About Pipelock

    Pipelock is an open-source agent firewall built by PipeLab. It sits between an AI agent and the network as a local proxy, scanning mediated HTTP, WebSocket, MCP and A2A traffic for secret exfiltration, prompt injection, SSRF and tool poisoning. It ships as a single Go binary with an Apache 2.0 core, and the site lists v3.7.1 as the current release.

    What It Is

    Pipelock is an egress and MCP security control for agents that hold credentials and have shell or API access. The agent keeps its keys, and Pipelock inspects the traffic crossing the boundary. Enforcement applies to traffic routed through its proxy, MCP wrapper, sandbox or containment setup; a direct network path around it must be prevented by routing or containment.

    It works in three modes: strict (allowlist-only), balanced (the default) and audit. Mediated enforcement paths fail closed, so timeouts, parse errors and malformed messages are blocked rather than passed.

    How the Scanning Works

    URL scanning follows a fixed order: preflight checks, destination policy, credential and DLP checks, then DNS resolution, so secrets are caught before a DNS lookup occurs. The docs list 65 built-in DLP patterns and 34 response-injection patterns with multi-pass normalization.

    For MCP, it wraps stdio servers or bridges to HTTP servers, scans tool arguments and responses, detects tool poisoning and mid-session description changes, and applies tool policy and chain detection. Optional controls include a seven-source kill switch, adaptive per-session enforcement, a Filesystem Sentinel, Browser Shield, and process containment using Landlock, seccomp and network namespaces on Linux or sandbox-exec on macOS.

    Evidence and Receipts

    Pipelock emits Ed25519-signed action receipts and a hash-chained flight recorder. Receipts can be checked offline with pipelock verify-receipt, and a free evidence viewer renders a recorded session as an HTML report. pipelock assess combines attack simulation, config scoring, install verification and discovery into signed evidence bundles. Verifier implementations exist in Go, TypeScript, Rust and Python.

    Setup Path

    Install from source with Go, or via Docker, Helm (OCI), Homebrew or a release binary. pipelock claude setup wires Claude Code through the proxy, and pipelock mcp proxy wraps a stdio MCP server. Integration guides cover Cursor, VS Code, JetBrains, Zed, Codex CLI, Cline, OpenCode, LangGraph, CrewAI, Google ADK and AutoGen. A GitHub Action can scan diffs in CI. Paid Pro and Enterprise tiers add multi-agent profiles, a dashboard and a fleet control plane.

    Pipelock - 1

    Community Discussions

    Be the first to start a conversation about Pipelock

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

    Pricing

    OPEN SOURCE

    Community (Open Source Core)

    Free Apache-2.0 licensed core firewall; run on as many agents as you want.

    • Apache 2.0 core
    • Scanning, detection, and enforcement controls in the open-source binary
    • Single Go binary
    • HTTP, MCP, WebSocket traffic
    • Signed action receipts

    Capabilities

    Key Features

    • Ordered URL scanner pipeline with DLP before DNS
    • 65 built-in DLP patterns with encoding-aware inspection
    • Response scanning for prompt injection
    • MCP proxy with tool poisoning and rug-pull detection
    • MCP tool policy and tool-call chain detection
    • WebSocket and A2A traffic inspection
    • SSRF and DNS rebinding protection
    • Signed action receipts verifiable offline
    • Hash-chained flight recorder
    • Seven-source kill switch
    • Adaptive per-session enforcement
    • Process sandbox and host containment
    • Browser Shield and media policy
    • Canary tokens and skill-file scanning
    • Signed community rule bundles
    • Security assessment with signed evidence bundles
    • GitHub Action for diff scanning
    • Prometheus metrics and SIEM event emission

    Integrations

    Claude Code
    Cursor
    VS Code
    JetBrains
    Zed
    Codex CLI
    OpenClaw
    Continue
    Cline
    OpenCode
    LangGraph
    CrewAI
    Google ADK
    AutoGen
    OpenAI Agents SDK
    Docker
    Kubernetes
    Helm
    GitHub Actions
    Prometheus
    Grafana
    Splunk
    Elastic
    Microsoft Sentinel
    OpenTelemetry
    API Available
    View Docs

    Demo Video

    Pipelock Demo Video
    Watch on YouTube

    Ratings & Reviews

    No ratings yet

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

    Rate other tools you’ve used

    Developer

    PipeLab

    PipeLab builds Pipelock and maintains open agent-security projects. It also publishes Agent Egress Bench, Pipelock Rules and the Agent Evidence Level draft standard. The company says it started as a homelab project to secure AI agents with shell access and API keys.

    Founded 2024
    Atlanta, GA
    Read more about PipeLab
    WebsiteGitHubLinkedInX / Twitter
    1 tool in directory

    Similar Tools

    Eidolon AI icon

    Eidolon AI

    An open-source AI Agent Server and SDK for building and deploying enterprise-grade agentic applications with declarative YAML configuration and Kubernetes support.

    Eve icon

    Eve

    Eve is Vercel's open-source agent framework that lets developers build, run, and scale production AI agents using a file-based convention—no infrastructure assembly required.

    LiveKit icon

    LiveKit

    Open source framework and cloud platform for building, deploying, and scaling voice, video, and physical AI agents in production.

    Browse all tools

    Related Topics

    MCP Tools

    Tools built with the Model Context Protocol for specific tasks.

    165 tools

    AI Infrastructure

    Infrastructure designed for deploying and running AI models.

    589 tools

    Network Security

    Tools for securing networks, including VPNs and firewalls.

    25 tools
    Browse all topics
    Back to all toolsSuggest an edit
    ratings
    discussions
    14views