Addy Osmani: Running Agents in Old Codebases — Brownfield Agentic Engineering
rseroter · x · 2026-09-18
Chrome engineering leader Addy Osmani published a deep dive on Brownfield Agentic Engineering — making AI coding agents safe in legacy codebases:
- The problem: old repos are no longer a complete description of system behavior; institutional knowledge, duct tape, and legacy services live outside the tree. Unsupervised agents produce code that 'works' but with wrong design and brittle tests.
- Goal: make hidden constraints visible and cheap changes trustworthy.
- Cautionary tale: Teleport spent a quarter with 13 engineers building a multi-agent vulnerability harness (skeptic/judge/summarizer agents) — it lost to someone opening one file and typing a simple prompt.
- Some argue any code not authored decision-by-decision is already a brownfield project.
rseroter noted every consulting shop now has playbooks for this, constantly updated.
More from coding & agent
- Anthropic engineer: the longer your agent runs, the costlier its mistakes — build verification in — Roger_M_Taylor · 2026-09-18
- Dev builds LLM-free chatbot wired to tools, sparking debate on what LLMs add beyond speed — jasonkneen · 2026-09-18
- claude-code-best-practice: a GitHub index mapping Claude Code concepts to real setups — Roger_M_Taylor · 2026-09-18
- Free 1-hour Grok Bot agents course: from one prompt to a 24/7 autonomous system — Roger_M_Taylor · 2026-09-18
- Fine-tune 500+ open source models for free in Google Colab with Unsloth Studio — Roger_M_Taylor · 2026-09-18
- Loops vs. Graphs: the loop lives inside a node, the graph lives between them — Roger_M_Taylor · 2026-09-18