List AI Prompt-Guard Presets

Endpoint

GET /apiops/settings/ai-prompt-guard-presets/

Authentication

Requires a Personal API Access Token with admin privileges.

Header

Authorization: Bearer YOUR_TOKEN

Request

Headers

Header	Value	Required
Authorization	Bearer {token}	Yes

Path Parameters

None.

Query Parameters

Parameter	Type	Required	Default	Description
projectId	string	No	admin	Scope project id. Omit for the admin/global scope, where built-in and admin-owned presets live.

Response

Success Response (200 OK)

{
  "status": "SUCCESS",
  "resultList": [
    {
      "id": "665f3b2d8c4f5b0013cd45ef",
      "projectId": "admin",
      "name": "BUILTIN_APG_DAN",
      "ruleValue": "(?i)ignore (all|previous) instructions",
      "action": "BLOCK",
      "category": "jailbreak",
      "description": "Classic DAN / instruction-override jailbreak attempt",
      "enabled": true,
      "builtIn": true,
      "owaspLlmId": "LLM01",
      "mitreAtlasTechnique": "AML.T0054",
      "packageVersion": "1"
    },
    {
      "id": "665f3b2d8c4f5b0013cd77aa",
      "projectId": "admin",
      "name": "block-internal-hostnames",
      "ruleValue": "(?i)internal\\.corp\\.local",
      "action": "FLAG",
      "category": "system-leak",
      "description": "Flags mentions of internal hostnames",
      "enabled": true,
      "builtIn": false
    }
  ],
  "resultCount": 2
}

Response Fields

Field	Type	Description
status	string	Response status: SUCCESS or FAILURE
resultList	array	List of prompt-guard preset objects
resultCount	integer	Total number of presets returned

Prompt-Guard Preset Object Fields

Field	Type	Description
id	string	Server-assigned identifier (read-only)
projectId	string	Scope the preset belongs to (admin for global)
name	string	Unique preset name (e.g. BUILTIN_APG_DAN)
ruleValue	string	Java regex pattern matched against prompt content
action	string	Action applied on match: BLOCK, FLAG, MASK
category	string	Optional category label (e.g. jailbreak, injection, system-leak)
description	string	Optional description for display
enabled	boolean	Whether the preset is active
builtIn	boolean	true for Mongock-seeded rows, false for user-created — the field itself is server-owned, but the row is not otherwise read-only
owaspLlmId	string	Optional OWASP GenAI LLM Top 10 category id (e.g. LLM01, LLM08). Read+write — a custom preset may set its own value; built-in rows are stamped by the signature-pack migration.
mitreAtlasTechnique	string	Optional MITRE ATLAS technique id (e.g. AML.T0054). Same read+write contract as owaspLlmId.
packageVersion	string	Read-only, system-managed: the signature-pack version that stamped this built-in row. null for custom presets and for built-in rows not yet touched by the signature-pack migration.

EnumStatus

• SUCCESS - Operation successful
• FAILURE - Operation failed

Error Response (400 Bad Request)

Returned when the caller is authenticated but lacks the ADMIN role.

{
  "status": "FAILURE",
  "resultMessage": "Unauthorized! Only users with ADMIN role can manage AI presets!"
}

Error Response (401 Unauthorized)

{
  "status": "FAILURE",
  "resultMessage": "Token is not valid!"
}

cURL Example

Default (admin/global scope)

curl -X GET \
  "https://demo.apinizer.com/apiops/settings/ai-prompt-guard-presets/" \
  -H "Authorization: Bearer YOUR_TOKEN"

Scoped to a project

curl -X GET \
  "https://demo.apinizer.com/apiops/settings/ai-prompt-guard-presets/?projectId=MyProject" \
  -H "Authorization: Bearer YOUR_TOKEN"

Notes and Warnings

• Admin Only:
  • Only sysAdmin users (or users with the ADMIN role) can list AI prompt-guard presets
  • A project-scoped AI_DEVELOPMENT token is not sufficient
• Built-in Presets:
  • Rows with builtIn: true come from the seeded OWASP LLM Top 10 / MITRE ATLAS signature pack. They are editable and deletable like any other row — see Built-in Presets Are Editable.
• Signature Pack Labels:
  • owaspLlmId/mitreAtlasTechnique are display/reporting metadata only — they never affect runtime matching. Today's built-in pack maps two categories: LLM01 (Prompt Injection) for jailbreak/injection/role-override/safety-bypass patterns, LLM08 (Hidden Context Exposure) for system-prompt-leak patterns.
  • packageVersion is read-only and only ever set by the signature-pack migration — sending it on write has no effect
• Action Semantics:
  • BLOCK rejects the traffic (403); FLAG lets it pass and records a guardrail hit; MASK redacts the matched region
• Scope:
  • The default scope is admin/global; pass projectId to list presets for a specific project
• No Secret Fields:
  • Prompt-guard presets carry no secret (@SecretData) fields, so no values are masked

Related Documentation

• Get AI Prompt-Guard Preset - Get a single preset by name
• Create AI Prompt-Guard Preset - Create or update a preset
• AI Prompt-Guard Presets API - Resource overview