Overview
Keystone is an AI engineer that integrates with your team, comprehensively understands your product, and maintains your codebase like a human would.
Key Features:
- Contextual understanding
- Autonomous software maintenance
- Integration with existing teams
- Monitoring and issue reproduction
- Proposing fixes and validating changes
Use Cases:
- Enhancing software development efficiency
- Reducing time spent on debugging
- Improving team collaboration
- Streamlining code maintenance
- Facilitating production monitoring
Benefits:
- Increased engineering productivity
- Less reliance on manual code reviews
- Faster issue resolution
- Enhanced understanding of user behavior
- Reduction of software failures
Capabilities
- Creates and manages end-to-end test suites in the Keystone dashboard and via API
- Triggers test suite execution via REST API POST /v1/suites/{suite_id}/run
- Returns run status and results via GET /v1/runs/{run_id} and dashboard run pages
- Integrates as a GitHub App to run tests on pull requests and surface PR status checks
- Provides a GitHub Action (keystone-platform/test-action@v1) to run tests inside workflows
- Accepts staging_url from CI deployments (e.g., Vercel) and applies staging URL overrides
- Tags tests with branch origin, auto-promotes tests on PR merge, and clears dev overrides
- Publishes detailed failure reports with direct links to test results from PR checks/dashboard
- Supports scheduled execution and monitoring of promoted production tests with alerting
- Exposes workflow parameters (suite_id, wait_for_results, fail_on_error) for Actions usage
- Offers programmatic CI/CD examples for GitHub Actions, GitLab CI, Jenkins, and CircleCI
- Authenticates API requests using API keys obtained from the Keystone dashboard