David Tandoh's working playbooks — distilled, opinionated guides for building things well. Each playbook synthesizes primary sources into a checklist-driven reference I can actually execute against.
| Playbook | What it covers |
|---|---|
| Building AI Agents | When to build an agent, the ReAct loop, single-agent tool/context design, multi-agent orchestration, failure handling |
| Building Reliable Agents | Observability (OTel GenAI), token cost & economics, reliability engineering & evals, security/prompt injection, and the production blind spots |
- One playbook per Markdown file at the repo root.
- Every playbook cites its primary sources up top.
- Favor tables, checklists, and rules-of-thumb over prose — these are references, not essays.
- Update the index table above when adding a playbook.