Integrations · CI / Automation
Trust Scanner CI
Enforce AI trust language standards in every pull request.
Screenshots coming soon
Visual walkthrough of Trust Scanner CI in progress
Overview
The SecEng Trust Scanner CI adds trust-language scanning as an automated quality gate to your GitHub Actions workflows. Run it as a workflow step to scan changed files, PR descriptions, or documentation for AI security claim issues — and fail the build if high-risk language is detected. The CLI mode also works in GitLab CI, CircleCI, and any other CI environment that can run Node.js.
Features
- 01.
GitHub Actions integration
Add the trust scanner as a workflow step in minutes using the provided GitHub Actions action definition.
- 02.
PR description scanning
Automatically scan pull request descriptions for AI trust-language issues as part of your review process.
- 03.
File-level scanning
Scan changed files in the PR diff — docs, policies, configuration — to catch issues before they merge.
- 04.
Build gate on findings
Configure the scanner to exit with a non-zero status code when high-risk language is detected, blocking the merge.
- 05.
Evidence artifact output
Upload scan results as GitHub Actions artifacts for audit-trail evidence attached to the workflow run.
Install steps
Step 01
Add the Trust Scanner CI action to your `.github/workflows/` YAML file.
Step 02
Set `GITHUB_TOKEN` in your workflow environment for PR description access.
Step 03
Configure the `scan-paths` and `fail-on-risk-level` inputs in the action step.
Step 04
Push a commit or open a PR — the scanner runs and reports findings as a check.
Capabilities
Privacy architecture
Local first
Designed to keep the supported flow in the local or customer-controlled environment. What leaves the device or workspace depends on the feature and configuration.
Platform vendor
GitHub, Inc.
GitHub, Inc.'s package json format is mapped in the registry. The connector itself is not yet built.
Early access
Get the integration on the roadmap
Tell us which platform matters most and we'll prioritize it where it fits the current product surface.