Building the CyberSG Atlas Chatbot in Four Weeks

A small RAG chatbot for a public cybersecurity directory, the guardrails that had to live in code rather than in the prompt, and the retrieval bugs real questions found.

November 12, 2026 · 5 min

Supply Chain Hygiene for Claude Code Plugins

Pinning every external dependency in your dev environment to an immutable SHA, and what that looks like in practice for Claude Code plugins.

October 19, 2026 · 5 min

OpenROAD MCP: Let AI Close Your Timing

Introducing OpenROAD MCP, which lets AI assistants like Claude Code analyze timing and adjust constraints on a live OpenROAD session.

February 19, 2026 · 3 min

Building Robust AI Research Agents: A tale of five specialists

What it took to build a deep research workflow for VCs, from a single search wrapper to five specialized agents with hierarchical citations.

November 20, 2025 · 24 min

Multi-Agent Systems Are 23x Cheaper Than Deep Research Models with Higher Quality

Multi-agent systems won or tied in all 5 scientific domains with higher novelty scores and 10-30x lower cost than deep research models.

November 13, 2025 · 8 min