Hadeda

Hadeda

·4 min read

What Is Hadeda? A Security Gateway for AI Coding Agents

Hadeda is a free, open-source VPC-resident security gateway that sits between your AI coding tools and your internal systems. Here's how it works and why it exists.

Hadedasecurity gatewayAI agentsopen source

AI coding tools like Cursor, Claude Code, and Windsurf are now part of engineering teams worldwide. They connect to internal APIs, databases, and production systems — often with the same access as the engineers who launched them.

That's a problem.

Hadeda solves it.

What Hadeda Does

Hadeda sits between your AI coding tools and your internal systems. Think of it as a security bouncer for AI agents.

Every command, API call, and tool request passes through Hadeda before it reaches your infrastructure. Hadeda enforces your policies, authorizes actions, and audits everything — in real time, with sub-2ms overhead.

The Core Features

Command Filter

Hadeda inspects every command your AI agent sends. You define rules — block certain patterns, allow specific tools, restrict access to sensitive systems. Dangerous commands get caught before they reach your infrastructure.

Secret Scrubbing

API keys, tokens, database credentials, and PII are stripped from agent outputs in transit. Sensitive data never leaks through logs, responses, or agent memory.

Audit Logging

Every action produces a structured JSON log entry — timestamp, user, tool, command, policy decision. These logs are SIEM-compatible out of the box, ready for SOC2 and ISO27001 audits.

Rate Limiting

Token bucket rate limiting per user and per tool prevents runaway agents from flooding your systems. Set limits that match your infrastructure capacity.

mTLS & Authentication

Gateway identity verification and API key validation ensure only authorized agents can pass through.

Why VPC-Resident Matters

Hadeda runs inside your network. Your data never leaves your environment.

This means:

  • No latency from cloud proxies
  • Full control over deployment and updates
  • No data residency concerns
  • Sub-2ms filter overhead

How It Works

Getting started takes under 10 minutes:

  • 1.Install dependencies
  • 2.Copy and configure your policy file
  • 3.Start the gateway
  • 4.Point your AI tools at Hadeda instead of your internal systems
  • That's it. Your agents now have guardrails.

    Who It's For

    Hadeda is built for two audiences:

    Engineers who want to use AI coding tools without being blocked by security teams. Hadeda gives you the paper trail that gets tools approved.

    Security and Compliance teams who need visibility into what AI agents are doing. Hadeda provides enforced policies and audit logs ready for SOC2 or ISO27001.

    Open Source and Free

    Hadeda is free and open source. Full core features, self-hosted, unlimited installs. No per-token surprises, no hidden costs.

    The paid tier adds a policy dashboard, email support, and additional integrations — but the core product is free forever.

    Why It's Called Hadeda

    The hadeda ibis is a bird found across southern Africa. It's loud, it's present everywhere, and it announces itself before anything happens. Like a good security gateway — it makes sure you know what's coming through.

    Getting Started

    Install Hadeda from GitHub. Configure your policies. Start protecting your AI coding workflow today.

    Every powerful system deserves guardrails.

    Your AI tools have unrestricted access. Now what?

    Hadeda is a free, open-source security gateway that sits between your AI coding tools and your internal systems. Filter commands, authorize actions, enforce policies, and audit everything — inside your network.