Connect Liongard to Conduit

This walkthrough connects Liongard (automated documentation) to Conduit. Liongard needs three values: your instance name plus an Access Key ID / Access Key Secret pair. Conduit combines the ID and secret into the single API key Liongard’s API expects, and sends every request to your instance — so all three values have to match.

Menu labels can change

The Liongard-portal steps below reflect Liongard’s access-key flow at time of writing; menu labels can shift between releases. If a label differs, see Liongard’s authentication guide.

Before you start

You need:

• A Conduit role that can manage connections — owner or admin on your organization.
• A Liongard administrator login.
• To know your instance name — look at the address bar when you’re signed in to Liongard: https://<instance>.app.liongard.com. The first label (for example us7 or eu1) is your instance name. You can type that label, or paste the host (eu1.app.liongard.com) or the full URL — Conduit keeps the subdomain either way. Access keys only work against the instance they were created on.

Step 1 — Create an access key in Liongard

1. Sign in to your Liongard instance (https://<instance>.app.liongard.com).
2. Create an API access key following Liongard’s authentication guide. Expected: an Access Key ID and an Access Key Secret, shown at creation.
3. Copy both values immediately. Liongard shows the secret only at creation — if you lose it, you’ll need to generate a new key pair.

Caution

Copy the Access Key Secret now and store it somewhere safe. Don’t paste it into chat, email, or a shared document — the next step moves it directly into Conduit, which encrypts it at rest. If you accidentally expose the secret, revoke the key pair in Liongard and generate a new one; the exposed pair becomes useless immediately.

Step 2 — Connect Liongard in Conduit

1. In Conduit, go to Connections (/org/catalog). Expected: the vendor catalog.
2. Find Liongard in the catalog (under Documentation) and click Connect. Expected: a credential form with three fields.
3. Fill the fields:
   • Instance Name (required) — your Liongard subdomain, e.g. us7 or eu1 (the placeholder shows us7). Pasting eu1.app.liongard.com or https://eu1.app.liongard.com is accepted and stored as that same subdomain.
   • Access Key ID (required) — from Step 1.
   • Access Key Secret (required) — from Step 1.
4. Click Connect. Expected: Conduit validates the key pair against your Liongard instance immediately and shows a green Connected status. A bad key or wrong instance name is rejected here, not on the first tool call.

The credential is stored at the organization scope, so every technician in your org shares this Liongard connection. (For per-technician keys, connect at the personal scope instead — see Vendor connections.)

Verify it worked

Connect an MCP client to Conduit — Claude Desktop, Claude Code, Cursor, or any MCP-compatible client (see Connecting an AI client) — and ask it to run a read-only Liongard tool, such as listing environments or recent detections. You should get real data back, and the call appears in your audit log with the correct user, vendor (liongard), and a successful status.

Troubleshooting

Symptom	Likely cause
Connect fails with “Invalid Liongard instance name. Use your Liongard subdomain, e.g. us7 or eu1.”	The value isn’t a Liongard subdomain. Use the first label from your Liongard address, for example us7 or eu1. A pasted Liongard host (eu1.app.liongard.com) or full URL (https://eu1.app.liongard.com) is fine; a different host is not.
Connect fails with “Liongard redirected the request unexpectedly. Check the instance name and try again.”	Liongard sent the check somewhere other than the instance you named. Confirm the subdomain (for example us7 or eu1) and try again.
Connect fails with “Invalid Liongard access key or instance name.”	Liongard rejected the request. Either the Access Key ID / Secret pair is wrong or revoked, or the instance name points at a different instance than the one the key was created on. Verify all three values together.
Connect fails with “Liongard returned HTTP NNN.”	Liongard responded with an unexpected status — often a transient outage or maintenance on your instance. Confirm you can sign in to the instance in a browser, then retry.
Tool calls return 403 “no credentials for vendor”	Liongard isn’t connected at a scope this user can resolve. Connect it at the organization scope, or confirm the user’s team has a key.
Vendor health shows yellow/degraded	Conduit’s health check sees Liongard responding slowly. Usually a transient Liongard-side load spike — recovers automatically.
Need to rotate the key	Generate a new access key pair in Liongard, then update the credential at /org/catalog.

Related

• Vendor connections — credential scopes, rotation, and the full vendor catalog.
• Security — how Conduit encrypts and resolves the credential you just stored.