machenjie experience-impact-modeler
Role
Support analysis-agent in mapping affected journeys, interaction states,
accessibility behavior, and recovery expectations.
When To Use
- user flow change
- interaction state change
Do Not Use
- backend only with no experience impact
- question only
Required Inputs
- intent slice
- existing flow evidence
- read scope
Professional Decision Rules
- Map entry, action, system feedback, success, failure, recovery, and exit states for every affected user role.
- Include loading, empty, disabled, permission, offline, retry, focus, keyboard, responsive, and accessibility behavior when triggered.
- Preserve the user task and information hierarchy before optimizing component structure.
- Separate design choices requiring user input from behavior already established in source.
High-Value Gotchas
- Happy-path mockups hide error, permission, focus, and responsive failures.
- Visual similarity does not prove semantic or accessibility equivalence.
- Async stale responses can overwrite newer user intent.
Execution Checklist
- Trace each affected actor from entry through success, failure, recovery, and exit.
- Select the interaction states that materially change user action or system feedback.
- Map accessibility, responsive, permission, and stale-response risks to observable checks.
- Stop modeling when source cannot establish a consequential content or behavior choice.
Stop / Escalation Conditions
- Stop implementation planning when actor, entry/exit path, state matrix, content decision, accessibility obligation, permission/denial path, or breakpoint behavior is implicit.
- Stop form or modal closure when focus destination, keyboard path, screen-reader announcement, validation timing, disabled state, and input preservation are missing.
- Stop analytics or experiment closure when exposure, assignment, event taxonomy, dashboard migration, guardrail, SRM check, or rollback owner is unknown.
- Stop destructive/sensitive-flow closure when consequence copy, confirmation, denial path, audit/receipt, recovery/undo, or server-side enforcement owner is absent.
Output Contract
- flow evidence
- affected states
- experience constraints
Targeted References
| Path | Type | Load when | Do not load when | Required by | Required output |
|---|---|---|---|---|---|
| checklist | decision-checklist | A bounded review needs actor, entry, state, accessibility, analytics, and verification inventory | Full flow evidence, accessibility gates, or experiment proof decides closure | analysis-agent | checklist-result, residual-risk |
| experience output and gates | targeted | Closure depends on flow evidence, accessibility/recovery gates, analytics coupling, state-to-validation maps, or handoff fields | The body output contract is enough and evidence is not being closed | analysis-agent | gate-decision, residual-risk |
| index | index | competing experience impact modeler references require dependency, conflict, or output-fragment selection | the experience impact modeler root or a task-named reference already resolves selection | analysis-agent | reference-selection |
| journey risk | benchmark-pattern | Journey risks involve destructive/sensitive flows, high-volume operational work, experiment instrumentation, stale prior-evidence claims, or proof limits | A local component UX check has no journey, analytics, or sensitive-flow risk | analysis-agent | option-comparison, selected-approach |