BlueArch
You are working with an agentic coding assistant and, twenty minutes in, you get that itch. It edited the same file for the fourth time. It ran the same failing test again. You typed “no, that’s stil…
Keep reading
See all posts →More from the journal.
Rust tools that give AI agents real code context
CodeGraph and AgentGateway show how AST-backed knowledge graphs replace grep-and-guess workflows for agentic coding.
Joel Proctor · 3 minRead ↗
The agent that wrote it shouldn't verify it
Two open-source tools show how separated executor and verifier agents catch what self-review misses.
Joel Proctor · 4 minRead ↗
Your Codebase as a Graph You Can Ask
Two open-source projects turning repos and agent memory into searchable, self-hosted knowledge graphs.
Joel Proctor · 1 minRead ↗
Subscribe on Substack
More from the BlueArch Journal.
One email a week. Field notes on AWS cost, tagging, and commitment strategy — straight from the people building the control plane.
Free forever·No spam·Unsubscribe anytime