Glossary

A contract that lets one piece of software use another piece of software.

Agentic development lifecycle. The workflow for using agents across planning, coding, testing, review, deployment, and operations while keeping human ownership.

Specific conditions that must be true before work is considered done.

The structure of a system: components, responsibilities, boundaries, and tradeoffs.

Concrete proof of work: code, diagram, test result, runbook, log, API contract, or demo.

The trusted server-side part of an application. It usually handles business logic, persistence, permissions, and integrations.

Temporary fast storage used to avoid repeating expensive work.

Continuous integration. Automated checks that run when code changes.

Continuous delivery or deployment. The process of preparing or releasing code to an environment.

Software that talks to a server. A browser app, mobile app, CLI, or another service can be a client.

A part of a system with a responsibility, inputs, outputs, and failure modes.

An agreement between components about input, output, behavior, and errors.

Durable storage for structured data.

The process of turning code or build artifacts into running software.

Forward deployed engineer. An engineer who works close to users and production to deliver practical systems under real constraints.

The user-facing part of an application. It handles interaction, display state, accessibility, and client-side flow.

The property that repeating the same operation has the same effect as doing it once.

A rule that should always remain true, such as “a sent response must have an approver.”

A controlled change to database schema or data.

The ability to understand a running system from logs, metrics, traces, and events.

A written review of an incident or failure that explains timeline, impact, root cause, fix, and prevention.

A place where work waits before a worker processes it.

Returning a system to a previous known-good version or state.

Operational instructions for starting, stopping, checking, debugging, and recovering a system.

Software development lifecycle. The process from problem discovery through design, implementation, testing, release, operation, and learning.

A running software component that exposes behavior to clients or other services.

What a system remembers.

Executable proof that a system behaves as expected under specific conditions.

A process that handles background jobs outside the direct user request path.