# Automation Builder

## Objective

Let customers wire a trigger, some conditions and an action so routine work runs itself.

A visual builder where a customer connects a trigger to conditions and actions, plus the runtime that executes it and a log of what happened.

## Before You Begin

This feature is being added to an application that already exists and already
works. Do not scaffold a new project, and do not assume a blank slate.

Inspect the codebase first and establish:

- The existing application structure and where code of this kind already lives.
- The framework and version in use.
- The existing design system — colours, spacing, typography, and component conventions.
- Existing UI components you can reuse instead of writing new ones.
- The existing database structure, if this feature needs to persist anything.
- The existing authentication and authorization system, if this feature is user-scoped.
- Dependencies already installed, so you don't add a library that duplicates one.
- The existing test setup and conventions.

Only start writing code once you understand the above. If the application
already implements part of this feature, extend it rather than replacing it.

## Implementation Instructions

1. Ship a fixed, curated set of triggers, conditions and actions drawn from what the app already does. A general-purpose scripting environment is not the minimum version, and offering one you cannot support is worse than offering ten steps that work.
2. Run automations through the app's existing background job system. If Background Job Dashboard or Scheduled Job Monitor exists, automation runs appear there rather than in a separate queue nobody watches.
3. Make the run log the centre of the feature, not an afterthought. A customer whose automation did nothing needs to see which run fired, what data entered each step, and where it stopped.
4. Record every action an automation performs in the app's existing audit trail, attributed to the automation and to the customer who owns it, so an unexplained change can be traced back.
5. Do not let an automation take an action the owning user could not take by hand. Evaluate permissions at run time against the owner, and pause the automation when those permissions are revoked.

## UI and UX Requirements

Match the application's existing design system exactly. Reuse its components,
spacing, and typography. This feature should look like it was always there.

## Responsive Requirements

Works on mobile, tablet, and desktop. Touch targets are large enough to hit on a
phone, and nothing overflows horizontally at 320px.

## Accessibility Requirements

- Fully keyboard navigable.
- Correct semantic elements and ARIA roles.
- Visible focus states.
- Meets WCAG AA contrast.
- Dynamic changes are announced to screen readers.
- Respects prefers-reduced-motion.

## Edge Cases

- One automation's action fires another's trigger, which fires the first again. Detect cycles when the automation is saved where the graph makes it visible, and cut off runaway chains at run time with a depth limit that stops the loop and reports it.
- Cap runs per period per customer, and when the cap is hit fail loudly. Silently dropping runs produces a customer who believes their automation is working for weeks.
- Every run needs a log showing the triggering record, the data as it entered each step, the condition results, and the outcome of the action. Without it, debugging is guesswork.
- A connected integration will lose its authorization. Pause the automation, tell the owner why, and hold or discard queued runs deliberately rather than retrying into a wall of authentication failures.
- Offer a dry run that evaluates the trigger and conditions against real data and reports what it would have done without doing it. Nobody should have to test an automation on live customer records.
- Editing an automation while runs are in flight must not corrupt them. Version the definition and let in-flight runs finish against the version they started on.
- A failed action needs a retry policy the customer can understand: how many times, how far apart, and what happens when it finally gives up.
- Deleting the record or field an automation depends on must disable it with a clear reason, not leave it failing quietly on every trigger.

## Testing

Exercise the feature end to end in the running application. Cover every edge case
above, then run the existing test suite and confirm nothing regressed.

## Acceptance Criteria

- [ ] A customer can build, enable, disable and delete an automation from a visual canvas.
- [ ] Runs execute through the app's existing background job infrastructure and are visible in its dashboard.
- [ ] Every run has a log showing input data at each step, condition results, and the final outcome.
- [ ] Loops are detected at save time and cut off at run time, with the customer told what happened.
- [ ] Run caps are enforced and hitting one surfaces a visible failure rather than dropped runs.
- [ ] An integration losing authorization pauses the automation and notifies the owner.
- [ ] A dry run reports intended actions without performing any of them.
- [ ] The feature matches the existing design system.
- [ ] No existing functionality is broken.

## Adaptation Rules

- Match the existing design system. Do not introduce a new colour palette,
  spacing scale, or component library.
- Reuse existing components and utilities wherever they fit.
- Follow the naming, file layout, and code style already present.
- Do not upgrade, replace, or remove existing dependencies to make this
  feature fit. Adapt the feature to the app, not the app to the feature.
- Do not break existing functionality. If a change is genuinely required in
  existing code, make the smallest one that works and say so.
- If something in these instructions conflicts with how the application is
  built, follow the application and explain the deviation.

## Final Verification

Before you report the work as done:

1. Re-read the acceptance criteria above and check each one against what you
   actually built.
2. Run the application and exercise the feature end to end.
3. Run the existing test suite and confirm you have broken nothing.
4. Check the feature on mobile, tablet, and desktop widths.
5. Check keyboard navigation and focus handling.
6. Summarize what changed: files added, files modified, and anything you
   deliberately did differently because of how this application is built.

If any acceptance criterion is unmet, fix it before reporting completion.