Ploba logo

Find the right AI tool for the job: apps, agents, APIs, MCP servers and datasets, explained plainly.

Get the AI Digest

Discover

  • Explore
  • Top Charts
  • Categories
  • Agents
  • MCP servers
  • Developer tools

Learn & community

  • Learn AI
  • AI Glossary
  • Blog
  • Hackathons
  • People
  • AI Digest

For makers

  • List your tool
  • Docs
  • Claim a listing
  • Advertise
  • Write for us

Company

  • About
  • Careers
  • Changelog
  • Trust
  • Security
  • Contact

© 2026 Ploba. All rights reserved.

  • Privacy
  • Terms
  • Cookies
  • Advertising
  • All policies
llms.txt
Ploba logo
DiscoverHackathonsPeopleLearnAI Digest
Sign inGet started
HomeTop ChartsExploreCategories
TypesWeb apps1,894Agents585Plugins416APIs376Mobile apps152Datasets85
AdvertisingYour product hereReach people browsing AI tools on Ploba.Advertise hereAdvertisingYour product hereReach people browsing AI tools on Ploba.Advertise here
  1. Explore
  2. /
  3. Developer Tools
  4. /
  5. MCP Airlock
Developer Tools
MA

MCP Airlock

by Shalimov04

MCP Airlock is a proxy placed between an AI agent and an MCP server. It adds per-caller tool permissions, dry runs, human confirmation for risky calls, and an audit trail.

Plugin·Contact for pricing

Visit registry.modelcontextprotocol.io(opens in a new tab)

About MCP Airlock

Key features of MCP Airlock

  • Per-caller allow list: unlisted tools are refused
  • Risk tiers L0 to L3 that can differ per environment
  • Forced dry runs and one-time confirmation for L2 calls
  • Blast radius limits per call and per principal over time
  • Response size cap and prompt injection marking
  • Audit records before and after each call, to JSONL and optional Postgres
  • Container image and Helm chart

Who MCP Airlock is for

  • Teams putting AI agents in front of production MCP servers
  • Platform and security engineers who need audit and approval controls
  • Developers testing agents against tools with destructive actions

How to use MCP Airlock

  1. 1

    Step 1: Install with uvx mcp-airlock --help or pip install mcp-airlock, adding extras such as postgres, otlp or regex if needed.

  2. 2

    Step 2: Write a policy.yaml that lists the tools and their risk tiers per environment.

  3. 3

    Step 3: Start the proxy with --policy, --upstream and --env, setting AIRLOCK_TRUST_PRINCIPAL_HEADER=1 on a laptop or using a JWT elsewhere.

  4. 4

    Step 4: Point your agent at the proxy instead of the server; docs/clients.md covers Claude Code and Cursor.

  5. 5

    Step 5: Check the audit file and run airlock-policy lint on your policy.

MCP Airlock FAQ

What happens to a risky call?

An L2 call runs first as a dry run and comes back as input_required with a signed requestState. After a person approves, the agent repeats it and it executes once.

How does it know who is calling?

From a JWT bearer token or an X-Airlock-Principal header from a trusted gateway, never from the request body. With no principal, the call is refused.

Can I run it in a container?

Yes. The image ghcr.io/shalimov04/mcp-airlock includes all three extras and runs as a non-root user.

Does it change what the agent sees?

It marks suspected injection phrases but does not change the content. It does cut responses to the output cap and hides tools the policy does not list.

Technical details

Requirements

  • An MCP-compatible client; provider credentials or local runtime dependencies may be required.

Details

Pricing
Contact for pricing
Type
Plugin
Platforms
  • CLI
  • MCP
Categories
Developer ToolsSecurityAutomation
Added
Sep 17, 2026
ShPublisherShalimov04

Built from public sources

We put this page together from MCP Airlock’s own website and docs so you can compare it with other tools. Ploba isn’t affiliated with MCP Airlock.

Work at MCP Airlock? Claim the listing to manage it yourself.

Something wrong here?

Related tools

See more
TATrueNAS AIOps MCPAIOps Tools · Contact for pricingTrueNAS AIOps MCP lets an AI agent run TrueNAS SCALE storage operations, with an audit log, budget guard and undo records on every call. It is a community project, not affiliated with iXsystems.Developer Tools · Contact for pricingIronClaw MCP iconIronClaw MCPIronSecCo · Contact for pricingRun untrusted agent shell commands inside a gVisor sandbox.Developer Tools · Contact for pricingPAProxy AIOps MCPAIOps Tools · Contact for pricingProxy AIops is an MCP server and CLI for audited operations on Traefik, Caddy and HAProxy reverse proxies. The same 28 tools work across all three.Developer Tools · Contact for pricing