PTENES
TRAINING 3 · PHASE 2

WebMCP Integrator

Turn user journeys and existing sites into coherent, recoverable WebMCP integrations that are ready to evolve.

Current siteobservable input WebMCP toolsexplicit contract Agent + personverifiable result
4

Integrator chapters

4

Available modules

24

Published topics

1

Lab

SHARED PROGRESS

Your journey continues across repositories

The same Formação 1 identifier records read items, questions, and notes. On GitHub Pages, the origin is shared; JSON keeps the portable fallback.

0 of 240%

Integrator course map

The top navigates by chapter. Within each chapter, modules follow the chapter.module numbering.

CHAPTER 2

Large-scale migration

Multi-page inventory, portfolio prioritization, and batch rollout.

Next
  1. 2.1Journey crawler
  2. 2.2Enterprise catalog
  3. 2.3Migration in waves
  4. 2.4Change governance

CHAPTER 3

Ecosystems

Design systems, CMS, e-commerce, and composite applications.

Next
  1. 3.1CMS and content
  2. 3.2Commerce and checkout
  3. 3.3Microfrontends
  4. 3.4Embedded applications

CHAPTER 4

Integrator deliverable

Quality, documentation, handoff, and applied project.

Next
  1. 4.1Integration contract
  2. 4.2Compatibility matrix
  3. 4.3Operational handoff
  4. 4.4Integrator final project

Chapter 1 · available modules

CROSS-CUTTING LAB

WebMCP Migration Studio

Paste an auditable catalog in JSON to receive a diagnosis of its contract, clarity, schema, lifecycle, risk, and example output. The lab does not register or execute tools.

Detailed content

1.13h

🧭 Tool design

Turn human journeys into small, distinct tools that are easy to choose.

0 of 60%
What it is:

A tool starts with a real user goal, not with every existing button or function in the code.

Why learn:

Mapping the start, decision, effect, and evidence prevents the integration from copying the interface without understanding its intent.

Key concepts:

observable objective · initial state · expected effect · final evidence

What it is:

Small tools fulfill a complete intent and do not compete for the same call.

Why learn:

When two descriptions seem to answer the same request, the LLM has to guess, and the correct selection rate drops.

Key concepts:

no overlap · precise verb · clear boundary · own output

What it is:

The input should request only the necessary data and represent closed alternatives with enums when possible.

Why learn:

Focused schemas reduce invented arguments, but actual validation is still required in execute and on the backend.

Key concepts:

minimum fields · useful enum · concrete description · real validation

What it is:

The agent may know the name spoken by the user before it knows the system's internal identifier.

Why learn:

Separating search, resolution, and action avoids requiring opaque IDs too early or accepting ambiguous names too late.

Key concepts:

name discovers · ID identifies · explicit resolution · ambiguity handled

What it is:

The catalog should track the route, selection, authentication, and current stage of the journey.

Why learn:

Fewer available tools mean less context and fewer impossible calls.

Key concepts:

contextual catalog · dynamic registration · visible session · cleanup

What it is:

Generic and overlapping names need to become a sequence that reveals intent and effect.

Why learn:

The implementation proves that the catalog guides the agent and also explains the flow to product, security, and QA.

Key concepts:

inventory · refactoring · selection test · documentation

View Complete
1.23h

🔄 State, errors, and recovery

Model dependencies, stale state, idempotency, timeouts, cancellation, and recovery.

0 of 60%
What it is:

Each step states what has already happened, which actions are available, and which transition is valid.

Why learn:

Without a state model, the agent may confirm before starting or repeat an effect that has already completed.

Key concepts:

explicit state · valid transition · precondition · effect

What it is:

Recoverable errors indicate a next action; terminal errors end the attempt with a verifiable reason.

Why learn:

Classification prevents blind loops and lets the agent explain real alternatives to the user.

Key concepts:

stable code · actionable message · next action · honest ending

What it is:

The state used to make a decision can change before execution, especially for job openings, prices, and permissions.

Why learn:

Revalidating at the time of the effect prevents a tool from using an old snapshot as current authorization.

Key concepts:

revalidation · state version · conflict · new reading

What it is:

An idempotency key lets you repeat a request without duplicating a subscription, charge, or submission.

Why learn:

Agents, networks, and users can repeat calls; the operation needs to distinguish a retry from a new effect.

Key concepts:

idempotency key · safe retry · previous result · single effect

What it is:

AbortSignal and timeout end work that has lost relevance or exceeded the operational window.

Why learn:

True cancellation preserves resources and prevents the interface from showing success after the person has given up.

Key concepts:

AbortSignal · timeout · visual cleanup · canceled result

What it is:

The integration must prove the happy path, invalid order, retry, cancellation, and concurrent changes.

Why learn:

A failure matrix turns resilience into verifiable behavior, not architectural intent.

Key concepts:

scenarios · evidence · reproduction · regression

View Complete
1.33h

🏗️ Migrating existing sites

Audit workflows, reuse existing logic, and introduce WebMCP as a progressive enhancement.

0 of 60%
What it is:

Map forms, handlers, services, endpoints, and states before proposing tools.

Why learn:

The audit reveals where the rule already exists and where the interface contains logic that needs to be separated.

Key concepts:

forms · JS functions · APIs · state

What it is:

Not every feature deserves to be migrated first; weigh frequency, value, stability, and reversibility.

Why learn:

The right scope delivers learning early without starting with a critical or poorly understood action.

Key concepts:

value · frequency · risk · stability

What it is:

The same domain function should serve both human clicks and WebMCP execution.

Why learn:

Sharing the logic prevents divergence between the visual flow and the agent’s path.

Key concepts:

common service · adapter UI · adapter tool · single rule

What it is:

The tool translates arguments, calls the existing service, updates the UI, and returns a structured response.

Why learn:

Small wrappers are easier to review and don't recreate the application inside the WebMCP registry.

Key concepts:

translation · existing service · synchronized UI · return

What it is:

The human workflow remains functional when WebMCP is unavailable or disabled.

Why learn:

An experimental technology is added as an additional capability, with monitoring and rollback.

Key concepts:

feature detection · fallback · rollout · rollback

What it is:

The dossier connects inventory, catalog, risks, implementation, tests, and publication plan.

Why learn:

The migration becomes reviewable by product, development, security, and operations.

Key concepts:

map · risk · functional proof · rollout

View Complete
1.43h

⚛️ Modern frameworks

Integrate WebMCP with SPAs, React, Angular, and Next.js with predictable registration and cleanup.

0 of 60%
What it is:

A framework-free implementation makes registration, dependencies, and cleanup explicit.

Why learn:

The reference reduces magic and helps diagnose problems hidden by hooks or dependency injection.

Key concepts:

pure function · explicit registration · AbortController · baseline

What it is:

The effect registers tools for the current state, and cleanup aborts the previous registrations.

Why learn:

Strict Mode, rerenders, and dependency changes must not duplicate names in the catalog.

Key concepts:

useEffect · dependencies · cleanup · Strict Mode

What it is:

A service receives dependencies, registers the tools, and exposes cleanup to the component or route.

Why learn:

Dependency injection keeps rules testable and separate from the visual lifecycle.

Key concepts:

service · DI · OnDestroy · testability

What it is:

document.modelContext only exists on the client and after the document is active.

Why learn:

Separating the client component, loading, and server API avoids accessing Document during SSR.

Key concepts:

use client · hydration · SSR · API route

What it is:

SPAs need to remove tools from the previous route and register only the capabilities of the new screen.

Why learn:

A stale catalog is stale UI: the agent sees an action that the person can no longer perform.

Key concepts:

router · unmount · current catalog · transition

What it is:

Different frameworks must fulfill the same observable contract for registration, execution, UI, and cleanup.

Why learn:

A shared matrix separates differences in ergonomics from actual differences in behavior.

Key concepts:

same contract · same tests · proven cleanup · compatibility

View Complete

INTEGRATOR CONTINUITY

Next: tool design

After the fundamentals, the same training continues in chapters 2, 3, and 4. Each chapter will have its own modules; the student remains in the Integrator role while expanding scale and governance.

View chapter 2 modules