Curated · no affiliate padding
Resources
The stuff worth your time, pulled from what actually works — not what ranks. Every link earns its place. This grows every week from the build log.
Docs
Canonical sources. When in doubt, these over a blog post.- Building Effective Agents Anthropic The reference mental model for agents vs workflows. Read this first.
- Effective context engineering for AI agents Anthropic How to feed an agent the right context — the skill that transfers across every tool.
- Claude Code best practices Anthropic Official playbook for the explore → plan → code → commit loop.
- Codex CLI features OpenAI Approval/sandbox modes, AGENTS.md, and headless codex exec — the canonical source.
- opencode docs SST The model-agnostic terminal agent. Good for learning that "the agent" and "the model" are separable.
- Windsurf Cascade docs Windsurf Agentic edits with checkpoints — teach rollback discipline.
- Build an MCP server MCP How agents get tools. The protocol behind every "connect my data" feature.
Course
Free, structured curricula you can work straight through.- AI Agents Course Hugging Face Free, hands-on, end-to-end agents curriculum.
- Building Code Agents with smolagents DeepLearning.AI Short, rigorous intro to code-writing agents.
- Evaluating AI Agents DeepLearning.AI The evals half nobody teaches — how to know your agent is actually good.
- MCP for Beginners Microsoft Open, multi-language MCP curriculum you can work straight through.
- Spec Kit — spec-driven development GitHub Spec → plan → agent implementation → verify. Works with any agent.
Video
The watch-list — creators who teach the real workflow.- Claude Code best practices (talk) Anthropic The workflow patterns straight from the team that built it.
- Claude Code vs Codex vs Cursor: honest comparison Theo (t3.gg) Grounded head-to-head so learners pick a tool without hype.
- The Only AI Coding Tools Worth Learning in 2026 Tech With Tim Cuts the noise for someone starting from zero.
- Cursor Crash Course Traversy Media The gentlest on-ramp for people who want a familiar VS Code UI.
- Built a $1M/y SaaS with Claude Code Nick Saraev Proof + workflow for the "actually ship it" promise.
Guide
Sharp, practical reads on the parts that trip people up.- How to write a good spec for AI agents Addy Osmani A good spec is the single biggest lever on agent output.
- CLAUDE.md & AGENTS.md guide Kemal Codes The context file is where beginners most often leave the agent blind.
- awesome-claude-code Community The community hub — skills, hooks, commands worth stealing.
- How I use Cursor + best tips Builder.io Real-world Cursor habits, not the marketing version.
- Deploy your vibe-coding projects for free Codecademy The deploy step where most first projects die.
- LLM evaluation: a beginner’s guide EvidentlyAI Plain-English intro to measuring AI output.
Pitfalls
What goes wrong, named — so you see it coming.- Why Vibe Coding Fails and How to Fix It DAPLab, Columbia The failure modes, named — the raw material for the anti-patterns lessons.
- 10 Things I Wish I Knew Before Vibe Coding Jason Lemkin / SaaStr Founder-level real talk on the 70% production gap.
Free · the newsletter
New resources, every week
The build log is where new links come from. Get it in your inbox plus the Mirror System reference agent.