Notes from the work
Build the thing.
Explain the hard part.
Practical field notes about AI agents and modern software systems—tested against real constraints, written so you can make a better next decision.
- Evidence
- Working code, observable results, and primary sources.
- Judgment
- Tradeoffs, failure modes, and when not to use the technique.
- Freshness
- Every tutorial shows when it was last reviewed and tested.
The field notes
Useful past the quickstart.
10 pieces, built to get you from promising demo to dependable system.
- Read the field note
AI agents · Multi-agent systems · Architecture
Agent Swarms: When Multiple AI Agents Help—and When They Make Things Worse
A practical decision framework for choosing a multi-agent architecture, plus a minimal coordinator that keeps concurrency, budgets, and failure handling explicit.
- Read the field note
Multi-agent systems · Performance · TypeScript
Make Agent Swarms Faster and Cheaper with Bounded Parallelism
Use dependency graphs, a tested worker pool, cancellation, deduplication, and budgets to reduce multi-agent latency without creating rate-limit and cost failures.
- Read the field note
AI agent evaluation · Continuous integration · AI testing
AI Agent Evals in CI: A Release Gate for Prompts, Models, and Tools
Build a provider-neutral agent evaluation suite with deterministic checks, repeated trials, safety cases, uncertainty, and a CI policy that catches regressions.
- Read the field note
AI agent memory · PostgreSQL · pgvector
AI Agent Memory with Postgres and pgvector: Schema, Retrieval, and Forgetting
Design tenant-safe agent memory with provenance, hybrid search, retention, deletion, and pgvector indexing that does not mistake every transcript for knowledge.
- Read the field note
Model Context Protocol · TypeScript · AI agents
Build an MCP Client in TypeScript: Discovery, Tool Calls, and Trust Boundaries
Connect to a local MCP server, verify capability discovery, call a typed tool, and handle remote transports without confusing discovery with authorization.
- Read the field note
Model Context Protocol · TypeScript · AI agents
Build a Safe MCP Server in TypeScript: Tools, Validation, and Transports
Build and test a real MCP stdio server, then make the transport, validation, error, and authorization decisions required for a remote deployment.
- Read the field note
Durable execution · AI agents · Workflow orchestration
Durable AI Agents: Retries, Idempotency, Human Approval, and Resume
Design long-running agent workflows that survive crashes, retry only safe work, bind approvals to exact actions, and reconcile ambiguous side effects.
- Read the field note
Model Context Protocol · Agent2Agent · AI architecture
MCP vs. A2A: Which Agent Protocol Do You Need?
A practical architecture guide to using MCP for tools and context, A2A for peer-agent work, and both without collapsing their trust boundaries.
- Read the field note
Browser agents · Playwright · AI security
Secure Browser Agents with Playwright: Isolation, Network Policy, and Approval
Build a browser-worker boundary that treats webpages as hostile, isolates credentials, blocks unapproved network destinations, and requires approval for consequential actions.
- Read the field note
OpenTelemetry · AI observability · Multi-agent systems
Trace AI Agents with OpenTelemetry: Spans, Tokens, Tools, and Redaction
Instrument agent, model, retrieval, and tool work in one trace using current OpenTelemetry GenAI conventions without exporting prompts or secrets by accident.