Live

GuardEntry × ServiceNow

Integrate GuardEntry with ServiceNow Flow Designer so every AI agent action is evaluated against your compliance policy before it executes. Setup takes about 10 minutes.

Prerequisites

•ServiceNow instance (Tokyo or later) with Integration Hub licensed
•A GuardEntry API key — get one free at guardentry.ai/connect
•Admin role or flow_designer + import_set_admin roles

Import the GuardEntry Spoke

1. In ServiceNow, go to System Update Sets → Retrieved Update Sets → Import Update Set from XML
2. Upload GuardEntry_Spoke.xml and click Upload
3. Open the imported update set → click Preview Update Set → resolve any conflicts → click Commit Update Set

The spoke adds two Flow Designer actions to your instance:

• GuardEntry: Evaluate Agent Action — call before any agent action
• GuardEntry: Log Policy Event — log custom events to the audit trail

Configure the Connection Alias

The Connection Alias stores your GuardEntry API key and base URL.

1. Go to Connections & Credentials → Connection & Credential Aliases
2. Find GuardEntry Connection (installed by the spoke) → click to open
3. Under Connections, click New:

Field	Value
Name	`GuardEntry Production`
Credential	create new — see below
Connection URL	`https://app.guardentry.ai/api/v1`

4. To create the Credential: click New Credential:

Field	Value
Type	API Key Credentials
Name	`GuardEntry API Key`
Request Location	`Header`
Request Parameter Name	`Authorization`
Request Parameter Value	`Bearer fgrc_k1_your_key_here`

5. Click Save on both the credential and connection.

Add to your Flow Designer flow

1. Open Flow Designer → open the flow where your AI agent runs
2. Click Add Action → search for GuardEntry
3. Select GuardEntry: Evaluate Agent Action and drag it before your first agent action step
4. Configure inputs:

Input	Value	Notes
`content`	Description of the action being taken	e.g. `"Update user record: " + trigger.fields.description`
`agent_id`	Your agent/flow identifier	Optional — helps with per-agent audit
`direction`	`ingress`	Always `ingress` for pre-action evaluation

5. Add a Condition after the GuardEntry action:

• If decision equals block → add Error step, set message to the reasoning output
• If decision equals require_approval → add a Request Approval step
• Otherwise → continue to the agent action

6. Click Save → Activate

Available outputs

The Evaluate Agent Action step returns:

Output	Type	Description
`decision`	String	`allow` \| `block` \| `require_approval` \| `uncertain`
`confidence`	Number	0–1 confidence score
`reasoning`	String	Human-readable explanation of the decision
`matchedRule`	String	Which policy rule triggered (if any)
`sessionId`	String	Session ID — include on egress call to link records

Troubleshooting

⚠️"Connection refused" or 401 errors — Check that the credential Authorization header value is Bearer fgrc_k1_... (with the Bearer prefix) and the Connection URL is exactly https://app.guardentry.ai/api/v1.

⚠️Update Set conflicts — The spoke only adds new records. Accept all incoming changes unless you have an existing GuardEntry Connection alias from a previous install — in that case, skip the alias conflict.

⚠️All decisions are allow — Your policy may be in Observability Mode. The 👁 observe badge in your GuardEntry dashboard means violations are logged but not blocked. Switch to enforcement from the policy settings page when ready.

Salesforce AgentForce Get API key