HomeAI Agency AcademyLesson 10
Module 03 · Lesson 10

Map the happy path and the hard paths

Design the normal journey and the exceptions before real customers discover them for you.

Last updated August 5, 202615–25 minutesFree AI agent course
What you will learn

Make a clear, safer operating decision.

You will be able to map the common route through an agent workflow and the specific conditions that require a different question, a pause, or a human handoff.

Why this matters

Good agent work is useful before it is impressive.

A demo usually follows the happy path. Production work includes incomplete details, conflicting records, angry people, outages, unsupported questions, and requests that look simple until the system acts on them.

Field note 10

Make the relationship visible.

AI AGENTS · FIELD NOTE 10Normal case + uncertainty + exceptionTHE PATHS01Clear request02Missing detail03Tool failure04Sensitive caseOriginal visual framework for Map the happy path and the hard paths.AI AGENTS · FIELD NOTE 10Normal case + uncertainty + exception01Clear request02Missing detail03Tool failure04Sensitive case
Use this framework to make map the happy path and the hard paths visible before you build.
Core concepts

The language that keeps the work clear.

Happy pathThe most common, low-risk sequence from trigger to completion.
Hard pathA foreseeable variation that changes what information, action, or owner is needed.
Uncertainty thresholdThe condition that causes the agent to ask, defer, or route rather than pretend certainty.
Safe failureA useful response that preserves context and gets the case to a person without making things worse.
The practical method

Work through the decision in order.

Draw the normal case

Show the minimal path from first signal to a completed outcome.

List missing-information cases

Identify which facts the agent can request and which absences require a person.

List action and tool failures

Decide what happens if a source cannot be read, an appointment is unavailable, or a system returns conflicting information.

Write the escalation experience

Specify the language, summary, owner, and expected next step a customer receives when the agent stops.

Worked example

A realistic, bounded implementation.

An agent helps customers reschedule a home-service appointment. The happy path verifies identity, checks approved availability, confirms the new slot, and records it.

A hard path appears when a customer has an emergency issue, when the CRM has two similar appointments, or when the proposed slot would breach an existing policy. The agent must not invent a resolution.

It says what it can do, captures the relevant details, keeps the customer informed, and creates a task for the dispatch owner with a concise summary. The handoff is part of the design, not an apology after a failure.

Build it in practice

Use this copyable working template.

Adapt it to the client’s evidence, policy, people, and tools. Do not treat placeholders as approved instructions.

Happy path: [steps]. Missing information: [question or handoff]. Conflicting data: [stop condition]. Tool failure: [fallback]. Sensitive case: [owner + message]. Customer promise during handoff: [sentence].
Spacebrain implementation

Put the operating system around the agent.

Use conditions, task assignment, conversation notes, availability checks, and routing automations to make hard-path handoffs visible and owned.

Practice

Before you move on

  • Map one normal case and four hard paths.
  • Run the map with an operator who handles exceptions today.
  • Write the exact customer-facing handoff sentence for each risky path.
  • The happy path is short and testable.
  • Common exceptions are mapped before launch.
  • The agent does not pretend a tool failure is success.
  • Every hard path has an owner and preserved context.

Build the operating layer around your agent.

Use the free Spacebrain workspace to keep contact context, handoffs, tasks, automation, and reporting together.

Start for free →