Knowledge Base Last verified: 2026-04-17 · next review in 303 days
A curated reference for developers working with AI coding tools. Every page carries a Last verified date; the CI job flags pages that drift past their refresh window.
New here? Start with the Cheat Sheet . The whole KB in one
skimmable page.
Cheat Sheet TL;DR one-pager: 3 tools, 3 frameworks, MCP + CLAUDE.md in 60 seconds, gotchas per tool.
Tool Comparison Matrix Side-by-side features and pricing for Claude Code, Cursor, Copilot, Windsurf, Codex, Aider, and more.
Framework Comparison Claude Agent SDK, OpenAI Agents SDK, Pydantic AI, LangGraph — when to pick each.
Pricing + Models Cross-vendor pricing and model-ID matrix, refreshed every 30 days.
New to agentic coding? Start here:
In-depth guides for the leading AI coding tools:
Building your own AI agents? Compare the options:
How to use these tools well:
Prompt Engineering — Patterns for coding agents
CLAUDE.md Guide — Project configuration best practices
Context Management — Use your context window well
Vibe Coding — When to let the agent drive, when to step in
Testing AI Code — Making AI-generated code safe to ship
Evaluation — Measuring whether your agent actually works
Observability — Tracing, logging, cost accounting
Permissions & Guardrails — Bounding the blast radius
Security — Prompt injection, exfiltration, PII, secrets
Cost Optimization — Prompt caching, batch API, model routing
Latency Optimization — Streaming, parallel tool use, regions
CI Integration — Running agents in CI, PR bots, budget guards
Glossary — Key terms defined