Agent Patterns · Learn
One hands-on course per concept — short lessons, real mechanics, instant-feedback quizzes.
A free, practitioner's course platform built on the agentpatterns.ai corpus. Start with Context Engineering; more courses are landing.
15 lessons Prompt Engineering Instructions that actually get followed — altitude, polarity, negative space, and the compliance ceiling. Start → 24 lessons Context Engineering Shape what your agent sees — attention, caching, compression, retrieval, and drift. The smallest set of high-signal tokens. Start → 15 lessons Tool Engineering Tools agents use well — schema design, token efficiency, error output, and result shaping. Start → 20 lessons Harness Engineering Design the agent loop itself — tools, hooks, sub-agents, permissions, and verification gates. Start → 9 lessons MCP Server Design Build Model Context Protocol servers agents can drive safely and discover on demand. Start → 12 lessons Verifying Agent Work Prove the agent did it right — deterministic guardrails, the verification ledger, golden journeys, outcome grading, and evals. Start → 10 lessons Observability See what your agent did — tracing, debugging, event sourcing, and evals. Start → 19 lessons Security The lethal trifecta, prompt injection, secrets handling, sandboxing, and egress control. Start → 11 lessons Agent Anti-Patterns The named failure modes and their fixes — infinite context, the yes-man agent, objective drift, and more. Diagnosis-first. Start → 9 lessons Multi-Agent Systems Orchestrate multiple agents — fan-out and synthesis, coordination contracts, verify-gated completion, and the failure taxonomy. Start → 12 lessons Agentic Workflows How to actually run agents day to day — plan-first loops, parallel fleets, headless CI, reversible setup, and eval-driven development. Start → 8 lessons GEO Generative Engine Optimization — getting your content found and cited by AI agents. Start → 8 lessons Loop Engineering Designing, controlling, and terminating the loops agents run in — so they converge on the goal instead of spinning, stalling, or burning budget. Start → 8 lessons Token Engineering The same result for fewer, cheaper tokens — the right model, the right token, the right cache, at the right time — without degrading output. Start → 11 lessons Agent-Assisted Code Review Reviewing — and authoring — code when agents write most of the diff. Tiered pipelines, the committee pattern, reproduce-before-report, and PRs built to be reviewed. Start → 11 lessons Mastering Claude Code Get the most out of Claude Code — plan mode, sub-agents, hooks, permissions, skills and plugins, worktrees, and scheduling. Start →