Ai Firewall Mcp

Name: Ai Firewall Mcp
Author: Akhilucky

by Akhiluckycommunityv1.0.1

An MCP server that analyzes incoming prompts for prompt injection, jailbreak attempts, exfiltration, and policy violations, then can sanitize or block suspicious content. It also exposes status and benchmarking tools for the firewall pipeline.

From the registry: Multi-agent LLM security layer detecting prompt injection and jailbreaks.

GitHub Repository Website

Install & configure

Prerequisites

optional
Install Python package runner (pipx) or package via pip
The README shows a portable stdio setup using pipx to run the published PyPI package. Alternatively, you can install the package with pip.
```
$ pip install pipx
```
https://pypi.org/project/ai-firewall-mcp/
optional
Optional Docker image
If you prefer the Docker-based MCP server instead of the Python package, pull the published image first.
```
$ docker pull akhilucky/ai-firewall-mcp:latest
```
https://hub.docker.com/r/akhilucky/ai-firewall-mcp

Install in your client

Paste into the AI you're chatting with — it will figure out the rest

Please install the `ai-firewall-mcp` MCP server into my current AI client (that's you).

Optional prerequisites:
- Install Python package runner (pipx) or package via pip — The README shows a portable stdio setup using pipx to run the published PyPI package. Alternatively, you can install the package with pip. Run: `pip install pipx` (https://pypi.org/project/ai-firewall-mcp/)
- Optional Docker image — If you prefer the Docker-based MCP server instead of the Python package, pull the published image first. Run: `docker pull akhilucky/ai-firewall-mcp:latest` (https://hub.docker.com/r/akhilucky/ai-firewall-mcp)

Canonical MCP server config (stdio transport):
- command: `pipx`
- args: ["run","ai-firewall-mcp"]
- optional environment variables:
  - `FIREWALL_MODE`: Firewall strictness mode: strict, moderate, or permissive. (example: `<strict|moderate|permissive>`)
  - `SIMILARITY_THRESHOLD`: Vector match threshold used for threat detection. (example: `<your-similarity-threshold>`)
  - `LOG_LEVEL`: Logging verbosity level. (example: `<your-log-level>`)

Note: The README indicates the MCP server uses stdio transport. The simplest portable install path is the published PyPI package via pipx. A Docker alternative is also documented with command `docker` and args `["run","-i","akhilucky/ai-firewall-mcp:latest"]`. The REST API described in the README is separate from the MCP server and should not be used as the MCP config.

Add this MCP server to my current client's config in the correct format for you. If you need secrets or credentials I haven't provided, ASK me — do not invent values or leave raw placeholders. After adding it, tell me how to verify the server is connected.

Note: The README indicates the MCP server uses stdio transport. The simplest portable install path is the published PyPI package via pipx. A Docker alternative is also documented with command `docker` and args `["run","-i","akhilucky/ai-firewall-mcp:latest"]`. The REST API described in the README is separate from the MCP server and should not be used as the MCP config.

Tools (5)

analyze_prompt

Analyze a prompt for injection, jailbreaks, exfiltration, and leakage

get_threat_breakdown

Return detailed per-signal scoring breakdown from the last analysis

sanitize_prompt

Clean a suspicious prompt while preserving legitimate content

get_firewall_status

Health check showing vector DB size, model status, and uptime

benchmark_firewall

Run the adversarial test suite and return detection statistics

Environment variables

FIREWALL_MODESets firewall mode to strict, moderate, or permissive

SIMILARITY_THRESHOLDControls the vector match threshold for detection sensitivity

LOG_LEVELSets logging verbosity

Packages

ai-firewall-mcp

pypiv1.0.1transport: stdio

Details

Language: Python
License: MIT
Released: Jun 9, 2026
Last commit: Jun 9, 2026
Registry name: io.github.Akhilucky/ai-firewall-mcp

Related Servers

Claude Flow

AI orchestration with hive-mind swarms, neural networks, and 87 MCP tools for enterprise dev.

68.0k

Codebase Memory Mcp

Persistent codebase knowledge graph. Survives session restarts and context compaction.

39.2k

Serena

A powerful toolkit for coding, providing semantic retrieval and editing capabilities.

28.1k