Skip to content

With passthroughOnError set to true on Azure Content Safety Content Moderation policy request doesn't proceed #1210

@IsuruGunarathne

Description

@IsuruGunarathne

Please select the area the issue is related to

Area/Policies (Policies, Policy Hub, Policy Engine etc)

Please select the aspect the issue is related to

Aspect/API (API backends, definitions, contracts, interfaces, OpenAPI)

Description

When passthroughOnError is set to true on Azure Content Safety Content Moderation policy, as per the description the request should proceed to the upstream. but it's blocked when the guardrail is triggered.

Image

Steps to Reproduce

  1. Deploy a self-hosted gateway (with Azure credentials added in config.toml) and a sample backend to log the request data on the same network.
  2. Configure an API proxy and point it to the backend you deployed.
  3. Add Azure Content Safety Content Moderation policy at API level and set Severity threshold for all categories to 1 and set passthroughOnError to true
  4. Deploy the API and send a request that passes the validation
  5. Notice the logs in the backend (request should reach the backend)
  6. Send a request that fails validation
  7. Notice that no new logs appear on the backend (when passthroughOnError is set to true requests should be sent to backend even if validation fails)

Severity Level of the Issue

Severity/Minor (Non-critical functionality. Can be fixed in future releases)

Environment Details (with versions)

No response

Metadata

Metadata

Assignees

No one assigned

    Labels

    Area/AIPoliciesIssues related to policies,guardrails in AI GatewayArea/PoliciesIssues related to any policy, policy hub, policy engine etcAspect/APIAPI definitions, contracts, OpenAPI, interfacesSeverity/MinorType/Bug

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions