# Solved By Code

> An independent AI research publication. Daily signal on frontier models, AI security, open-source releases, and where the field is heading.

Solved By Code is an independent AI research publication covering four beats: frontier models, AI security, open-source releases, and the long-term horizon. Every claim is sourced to primary material. Every open-source release is tested before review. Corrections are stamped in place with dated changelogs. No sponsored stories, ever.

## Core pages

- [Home](https://solvedbycode.ai/): Signal from the open AI frontier
- [The Lab](https://solvedbycode.ai/lab): Technical reviews of open-source AI tools — every review includes a merged PR contribution to the project.
- [Lab Charter](https://solvedbycode.ai/lab/about): The six-step editorial rulebook behind every Lab review: install, production use, architecture review, benchmark, contribution, verdict.
- [Learning Center](https://solvedbycode.ai/learn): In-depth guides on Claude Code, vibe coding, RAG, vector databases, and generative engine optimization.
- [Frontier (Blog)](https://solvedbycode.ai/blog): Daily briefings on frontier AI — model releases, lab moves, security, policy — plus long-form Horizon essays.
- [Masthead & Editorial Principles](https://solvedbycode.ai/portfolio): What this publication covers, publishing cadence, and the editorial rules behind every story.

## The Lab (open-source AI reviews)

- [OpenClaw Dashboard: What I Fixed Before I Could Trust It](https://solvedbycode.ai/lab/openclaw-dashboard-hmac-auth-and-costs-review): A deep review of the OpenClaw Dashboard — the open-source WebSocket control UI for the AI gateway OpenAI just acquired. Broken HMAC auth, no cost visibility, and two PRs that fix both.
- [How I Built a Visual Dashboard for OpenClaw: The AI Gateway OpenAI Just Acquired](https://solvedbycode.ai/lab/openclaw-dashboard-visual-ui-for-ai-gateway): The full story of building an open-source React/Next.js dashboard for OpenClaw — 12 pages, streaming chat, WhatsApp QR login, speech-to-text everywhere, and contributing to open source during an acquisition.
- [OpenCode Review: Benchmarking the 60k-Star Claude Code Alternative](https://solvedbycode.ai/lab/opencode-benchmark-review-january-2026): Comprehensive benchmark testing of OpenCode Zen models including GLM-4.7, Grok Code Fast, and Big Pickle against real-world coding tasks.
- [Chrome DevTools MCP: Finally, Your AI Can Actually See What It's Building](https://solvedbycode.ai/lab/chrome-devtools-mcp-giving-ai-eyes): The frustration of AI coding blind is over. Chrome DevTools MCP connects your AI assistant directly to Chrome, letting it see console errors, inspect the DOM, run performance traces, and debug in real-time.
- [Vibe Kanban: Revolutionary Orchestration or Overhyped Complexity? An Honest Review](https://solvedbycode.ai/lab/vibe-kanban-honest-review): A deep dive into Vibe Kanban - the tool promising to 10X your AI coding workflow. We examine what it actually does, who it's for, and whether the complexity is worth it.
- [Anthropic Skills Explained: Fancy Prompts or Something Actually Useful?](https://solvedbycode.ai/lab/anthropic-skills-complete-guide): A complete guide to Anthropic Skills - what they are, how they differ from agents and slash commands, real examples, and whether Anthropic is watching how you use them.

## Learning Center (guides)

- [Claude Code vs Cursor vs GPT-5.2 Codex: An Objective Analysis of Vibe Coding Tools in 2026](https://solvedbycode.ai/learn/claude-code-vs-cursor-vs-codex-vibe-coding-comparison-2026): A comprehensive, data-driven comparison of the three leading AI coding tools in 2026. With real benchmarks, security analysis, cost breakdowns, and honest recommendations for different developer types.
- [The Complete Guide to Generative Engine Optimization (GEO) in 2026](https://solvedbycode.ai/learn/complete-guide-generative-engine-optimization-geo-2026): Comprehensive guide covering how to make AI search engines cite your content, GEO vs SEO differences, market statistics, optimization strategies, and code examples.
- [The Complete Guide to RAG and Vector Databases in 2026](https://solvedbycode.ai/learn/complete-guide-rag-vector-databases-2026): A comprehensive 5000+ word guide covering everything you need to know about Retrieval-Augmented Generation and vector databases in 2026. Includes rankings, benchmarks, code examples, and industry-specific recommendations.
- [The Complete Claude Code Guide: Everything You Need to Know (Written by Claude Code)](https://solvedbycode.ai/learn/claude-code-complete-guide-2026): The definitive guide to Claude Code - hooks, plugins, slash commands, agents, permissions, and more. Written by Claude Code itself.

## Blog (field notes)

- [The Open-Weight Security Asymmetry](https://solvedbycode.ai/blog/open-weight-security-asymmetry) (2026-04-19): Closed-model AI safety was built on gatekeeping. Open-weight frontier models dismantle every lever. The defender ecosystem has not caught up, and the gap is getting wider — here is what practitioners running these models in production actually need to do about it.
- [The 2026 AI Inflection Point — By the Numbers](https://solvedbycode.ai/blog/2026-ai-inflection-point) (2026-04-19): A data-first look at why 2026 marks the shift from AI-as-novelty to AI-as-infrastructure: frontier-lab revenue curves, adoption cliffs, vibe coding going mainstream, and the agentic turn.
- [Claude Code 2.1: What 1,096 Commits Look Like](https://solvedbycode.ai/blog/claude-code-2-1-everything-new-january-2026) (2026-01-11): A deep dive into Claude Code 2.1 from someone using it daily
- [The GitHub Explosion: AI Coding Tools Reach Critical Mass](https://solvedbycode.ai/blog/github-ai-coding-explosion-january-2026) (2026-01-11): GitHub trending reveals the AI coding revolution