Blog
All Posts
Thoughts on AI, development, and building products.
How AgentLens reads a codebase
The architecture behind AgentLens: from a repo URL to a flow graph, scored agent opportunities, and a report. Offline-first, with a deterministic engine under the LLM.
November 6, 2026 · 4 min read
Nutlope/hallmark: Why It Matters
Hallmark is a tool that helps developers write cleaner, more human-like code by avoiding generic AI-generated patterns in tools like Claude Code, Cursor, and Codex.
July 15, 2026 · 5 min
openinterpreter/openinterpreter: Why It Matters
Open Interpreter is a lightweight AI coding agent designed to run on low-cost models, enabling users to execute and debug code locally with minimal resources.
July 15, 2026 · 5 min
Autonomous AI Agents in IDEs: The Future of Coding
Agents that understand entire codebases and solve architectural bugs autonomously, moving beyond autocomplete into true pair programming.
March 13, 2026 · 7 min
Q-Star v3: Self-Correcting AI That Eliminates Code Hallucinations
Neural-symbolic architecture pairs LLMs with deterministic solvers to validate code before it ships.
March 12, 2026 · 7 min
SWE-agent 2.0: Open-Source Autonomous Software Engineering
Turn language models into software engineering agents capable of resolving real-world GitHub issues.
March 11, 2026 · 8 min