AutoMerge is generally available

Last updated:

About Darkplane

Control room for AI-assisted development, tracking adoption, cost, and output quality across your engineering team.

Darkplane is the control room for AI-assisted development. It governs the left side of your software lifecycle: what AI builds, and feeds production context back to the next change.

Organizations advance through the six levels of agentic software engineering — from AI-assisted code completion to autonomous development loops. Most teams today operate at Level 3 (AI-generated, auto-reviewed), where AI code review runs on every change and humans still approve every merge. Darkplane provides the visibility and measurement infrastructure needed at each stage.

As coding accelerates with AI assistance, Darkplane shows what your coding agents build, what that costs, and whether AI-written changes hold up in production — linking every session to the PR it produced and every incident back to the change that caused it. By connecting to observability (logs, traces, and metrics) and Agent0, Darkplane helps close the loop from PR to production.

Darkplane provides two capabilities:

  • AI SDLC Insights provides visibility into the coding stage of your AI-assisted software lifecycle. It reads coding harness telemetry and your pull requests together, covering cost, adoption, productivity, tools and skills, and session-level detail, so you can make measurement-based decisions about how your team uses AI coding tools.
  • AutoMerge covers the review stage. AutoMerge scores each pull request and approves the safe ones. You start it in Dry run, where it only posts its decision as a comment until you switch it on.

More capabilities are on the way, extending this visibility across the rest of your development lifecycle.

Darkplane reads the tools your engineers already use and is OpenTelemetry-native, keeping your data under your control.

Further Reading