4 posts tagged with “AI-Assisted Development”
Monitoring AI-generated code is an observability problem, not a detection one. The four CI signals that shift after an assistant rollout, and what each means.
Measuring AI coding productivity does not need new metrics. Why acceptance rate misleads, why splitting DORA by author is a trap, and which metrics move first.
CI/CD for Copilot, Cursor, and Claude Code: how each assistant connects to a shared CI/CD data layer via MCP, the workflows it unlocks, and what each does best.
The AI assistant CI/CD context problem: coding assistants guess about CI state because they cannot see it. The MCP server gives them typed access to it.