Open-source AI tools, shipped by ThinkNext.

Open-source AI agent that takes a meeting recording, a tracker ticket, or a one-line prompt and ships a tested pull request. Self-hosted, your LLM key, your code never leaves your org.

Autonomous AI agent that turns labeled issues into tested pull requests, running in your existing CI. The hands-off cousin of Cascade for well-scoped backlog work. Local CLI ships today; GitHub Actions in v0.1.1; GitLab CI + Bitbucket in v0.2; Azure Pipelines in v0.3.

AI-driven functional testing agent. Point it at a URL and it explores the app, generates a focused Playwright test plan, runs it in fresh browser sessions, and reports failed scenarios, visual regressions, accessibility violations, and REST API contract findings. Self-healing on failed steps.

AI code/PR review agent. Point it at a pull request URL: it fetches the diff, asks your LLM to review it, and posts inline comments and a verdict back to the PR. Multi-VCS (GitHub + GitLab today; Bitbucket + Azure DevOps next). Self-hosted, BYO LLM key, source never leaves your network.

Production-to-dev bridge. Pipe JSON-lines log events in (Sentry, Datadog, OpenTelemetry, app logs): Beacon clusters them by stack signature, asks your LLM to summarize each cluster, and files or updates one issue per cluster on GitHub or GitLab. Optional --relay-label hands well-scoped issues to Relay for an autonomous fix PR.