About Liongard
Liongard is an MCP server for AI assistants and teams working with the Liongard platform. Its decision-tree tool design exposes inspectors, systems, metrics, alerts, users, groups, environments, and integrations, with support for API v1 and v2.
Key features
- 01Navigate to domain-specific tools on demand
- 02Work with inspectors and systems
- 03Access metrics and alerts
- 04Manage users, groups, and environments
- 05Use API v1 and v2
- 06View interactive detection cards
Use cases
- 01Inspect Liongard systems and their metrics
- 02Review alerts and detection details
- 03Browse environments, users, and integrations
MCP support
Exposes Liongard platform operations through MCP tools, using a navigation tool to reveal domain-specific tools as needed. Detection details can appear as read-only interactive cards in MCP Apps hosts.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
docker run -i --rm -e LIONGARD_API_KEY -e LIONGARD_INSTANCE -e MCP_TRANSPORT -e AUTH_MODE -e LOG_LEVEL ghcr.io/wyre-technology/liongard-mcp:v2.3.7LIONGARD_API_KEYSecretLiongard API access key (access key ID and secret combined per Liongard auth scheme)LIONGARD_INSTANCELiongard instance hostname (e.g. yourorg.app.liongard.com)MCP_TRANSPORTTransport mode for the server. Set to 'stdio' for local CLI use; the image defaults to 'http' for gateway hosting.AUTH_MODESecretCredential source: 'env' reads vars locally, 'gateway' expects header injection from the WYRE MCP Gateway.LOG_LEVELLog verbosity: debug, info, warn, errorLocal servers need a runtime on your machine (e.g. Node.js 18+). Config file locations differ by client; see your client’s docs.
ghcr.io/wyre-technology/liongard-mcp:v2.3.7stdioTools
Local servers expose tools only once running on your machine, and the maintainer has not provided a tool list.
Details
From the project READMELiongard MCP Server
An MCP (Model Context Protocol) server for interacting with the Liongard platform. Provides AI assistants with tools to manage inspectors, systems, metrics, alerts, users, and more through Liongard's API.
One-Click Deployment
Important
Before you click: this server depends on
@wyre-technology/node-liongard, which is hosted on the GitHub Packages npm registry. GitHub Packages has no anonymous access — even though the package is public, everynpm installneeds a token. The cloud builder runsnpm installfor you, so you must give it one, or the build fails withnpm error 401 Unauthorized ... npm.pkg.github.com.
- Create a GitHub Personal Access Token with the
read:packagesscope (classic token). Any GitHub account works — you do not need to be a member of thewyre-technologyorg to read its public packages.- Add it as a build variable when prompted by the deploy flow:
- Cloudflare Workers → set a build variable named
NODE_AUTH_TOKENto your PAT (Workers → Settings → Build → Variables and Secrets).- DigitalOcean App Platform → set an encrypted env var named
GITHUB_TOKENwith scope Build Time to your PAT (the Dockerfile reads it for the install).
Features
- Decision-tree tool architecture — starts with a navigation tool, then exposes domain-specific tools on demand
- 25 tools across inspectors, systems, metrics, alerts, users, groups, environments, and integrations
- Interactive detection card via MCP Apps (SEP-1865):
liongard_detections_getrenders as a read-only card in MCP Apps hosts (Claude Desktop/web) with severity/status badges and server-resolved system/environment names — neutral by default, brandable viawindow.__BRAND__injection orMCP_BRAND_*env vars (MCP_BRAND_NAME,MCP_BRAND_LOGO_URL,MCP_BRAND_PRIMARY_COLOR,MCP_BRAND_ACCENT_COLOR,MCP_BRAND_BG,MCP_BRAND_TEXT), no rebuild needed - Dual API version support (v1 and v2)
- 128 tests
Installation
This project depends on @wyre-technology/node-liongard, published to the
GitHub Packages npm registry, which requires a token even for public packages.
Authenticate once, then install:
# Authenticate npm to GitHub Packages (token needs the read:packages scope)
export NODE_AUTH_TOKEN=$(gh auth token) # or a PAT with read:packages
npm install
The repo's .npmrc already points the @wyre-technology scope at GitHub Packages and
reads the token from NODE_AUTH_TOKEN, so no further config is needed.
Configuration
Copy .env.example. Stdio and AUTH_MODE=env read the Liongard variables. Gateway mode does not.
| Variable | Required | Description |
|---|---|---|
LIONGARD_API_KEY | For stdio and AUTH_MODE=env | Liongard API key |
LIONGARD_INSTANCE | For stdio and AUTH_MODE=env | Liongard instance hostname (for example yourorg.app.liongard.com) |
MCP_TRANSPORT | No | stdio (default) or http |
AUTH_MODE | No | env (default) or gateway. Gateway mode requires X-Liongard-API-Key and X-Liongard-Instance on each /mcp request and never falls back to LIONGARD_*. |
CONDUIT_S2S_SECRET | Yes for HTTP | Service-to-service secret. The HTTP server logs an error and exits non-zero when this is empty or a known placeholder. Compose will not start without it. Never commit a real value. Generate one with openssl rand -hex 32. |
MCP_ALLOW_INSECURE_DEV | No | Set to 1 only for local development to start HTTP without CONDUIT_S2S_SECRET. The process logs a warning, does not enforce X-Gateway-S2S, and binds 127.0.0.1 only. A non-loopback MCP_HTTP_HOST refuses startup. A known placeholder secret is still refused. |
MCP_HTTP_HOST | No | Bind address when a real CONDUIT_S2S_SECRET is set. Unset keeps 0.0.0.0, which is what the container image sets, so container deploys stay on every interface inside the container. Without a secret the dev bypass binds 127.0.0.1 instead. |
MCP_HTTP_PORT | No | HTTP port. Defaults to 8080. |
/health stays unauthenticated and does not read credentials. The stdio transport does not use CONDUIT_S2S_SECRET.
HTTP authentication
With MCP_TRANSPORT=http and CONDUIT_S2S_SECRET set, every /mcp request must include X-Gateway-S2S. The value is t=<unix seconds>,v1=<hex>, where v1 is HMAC-SHA256 of the exact string t=<unix seconds> keyed by CONDUIT_S2S_SECRET (64 lowercase hex characters). The timestamp must be within 300 seconds of the server clock. A missing or invalid header returns 401. The secret is never written to logs.
import { createHmac } from "node:crypto";
const t = Math.floor(Date.now() / 1000);
const v1 = createHmac("sha256", process.env.CONDUIT_S2S_SECRET)
.update(`t=${t}`)
.digest("hex");
const xGatewayS2s = `t=${t},v1=${v1}`;
In AUTH_MODE=gateway, /mcp also requires X-Liongard-API-Key and X-Liongard-Instance. Those headers are the only credential source for that request. A request missing either header returns 401 and does not read LIONGARD_API_KEY or LIONGARD_INSTANCE from the environment.
Docker Compose
The checked-in Compose service is gateway mode. It publishes 127.0.0.1:8080:8080, requires CONDUIT_S2S_SECRET, and does not set LIONGARD_API_KEY, LIONGARD_INSTANCE, or env_file.
CONDUIT_S2S_SECRET=your-s2s-secret docker compose up --build
Usage
npm start
HTTP refuses to start until CONDUIT_S2S_SECRET is a real secret:
MCP_TRANSPORT=http CONDUIT_S2S_SECRET="$CONDUIT_S2S_SECRET" npm start
Local HTTP without a secret is development-only. It binds 127.0.0.1 and logs a warning:
MCP_TRANSPORT=http MCP_ALLOW_INSECURE_DEV=1 npm start
Development
npm run build # Build the project
npm run test # Run tests
npm run lint # Run linter
npm run typecheck # Type-check
License
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.