From FORSVN Dev
Maps multi-step in-product flows including screens, decisions, transitions, platform-native touchpoints, edge cases, and error states for feature design or audit.
How this skill is triggered — by the user, by Claude, or both
Slash command
/forsvn-dev:map-user-flow [feature or flow to map][feature or flow to map]This skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
*Design Step 2 of 2. Maps navigation paths, decisions, edge cases, platform-native touchpoints, and screen transitions into a validated flow (Mermaid + ASCII wireframes + per-surface mini-frames).*
agents/_template.mdagents/critic-agent.mdagents/diagram-agent.mdagents/edge-case-agent.mdagents/structure-agent.mdagents/validation-agent.mdagents/wireframe-agent.mdreferences/_shared/artifact-contract-template.mdreferences/_shared/before-starting-check.mdreferences/_shared/execution-policy.mdreferences/_shared/manifest-spec.mdreferences/_shared/mode-resolver.mdreferences/_shared/pre-dispatch-protocol.mdreferences/_shared/tool-redirect.mdreferences/anti-patterns.mdreferences/dispatch-protocol.mdreferences/examples/checkout-walkthrough.mdreferences/platform-touchpoints.mdreferences/playbook.mdreferences/pre-dispatch-prompts.mdDesign Step 2 of 2. Maps navigation paths, decisions, edge cases, platform-native touchpoints, and screen transitions into a validated flow (Mermaid + ASCII wireframes + per-surface mini-frames).
Core Question: "Can the user complete their goal without thinking — on every surface of every platform it ships on?"
Why, methodology, principles, history:
references/playbook.md[PLAYBOOK].
Design / redesign / audit a feature with multiple screens / decisions / states / platforms, or map a new platform-native surface (widget, Live Activity, watch app) into an existing flow.
Not for: brand (/create-brand) · API (/architect-system) · decomposition (/breakdown-tasks) · scoping (/discover) · single-page landing (/brief-landing-page).
Gate 1 (platforms+surfaces) fires before Pre-Dispatch — wireframe size, entry triggers, per-surface edge states all depend on it.
references/platform-touchpoints.md).research-icp.Full rationale + 13-platform list + critic-FAIL handling: references/procedures/gates-and-rubric.md [PROCEDURE].
Critic (agents/critic-agent.md) runs the full rubric. Non-negotiable PASS:
references/platform-touchpoints.md)Apply references/_shared/before-starting-check.md [PLAYBOOK]. budget: standard; --fast forces Single-Agent Fallback per references/_shared/mode-resolver.md [PROCEDURE]. Safety gates supersede --fast.
Session execution profile (single-vs-multi): inherit per references/_shared/execution-policy.md.
| Artifact | Source | Required? |
|---|---|---|
research/product-context.md | research-icp | Recommended (>30d → re-run) |
docs/forsvn/experience/{audience,technical,goals}.md | any | Recommended |
implementation-roadmap/canonical-paths.md | architect-system | Optional |
brand/{DESIGN,BRAND}.md | create-brand | Optional |
Dimensions: feature, role/persona, goal, platforms (explicit), surfaces per platform, primary surface, constraints (auth + min OS). Read order + Warm/Cold Start prompts (with platforms+surfaces gate) + write-back + brief-context contract: references/procedures/pre-dispatch.md · references/pre-dispatch-prompts.md [PROCEDURE].
docs/forsvn/artifacts/product-map-user-flow-<YYYY-MM-DD>-<flow-name>.md (flat v2; one file per flow; index auto-generated at ≥2 slugs). Lifecycle: pipeline.skill, version, date, status, stack: product, review_surface: html (FIRE preview while decision_state: pending), decision_state, flow_name, platforms, plus manifest-sync lifecycle, produced_by, provenance. v2 schema: references/_shared/artifact-contract-template.md.references/report-template.md [PROCEDURE].architect-system, breakdown-tasks, forsvn, review-work.Upstream tool-redirect.md.
Single route, full stack. Flows >15 screens auto-split via structure-agent sub-flow decomposition. Pipeline: Step 0 → L1 ∥ (structure ∥ edge-case) → Merge → L2a ∥ (diagram ∥ wireframe) → L2b seq (validation → critic). Critic FAIL re-dispatches (max 2 cycles).
Agent manifest, spawn mechanics, dispatch tables, Single-Agent Fallback: references/procedures/dispatch-mechanics.md · references/dispatch-protocol.md. Worked example: references/examples/checkout-walkthrough.md [EXAMPLE].
13-pattern catalog (happy-path-only edge cases, generic screen names, wireframe drift, "cross-platform" creep) + when-to-defer + critic-FAIL: references/anti-patterns.md [ANTI-PATTERN].
Prev: create-brand (optional). Next: architect-system → breakdown-tasks. Defer to: discover, create-brand, breakdown-tasks. Re-run on: feature changes, new research, usability failures.
research/product-context.md missing; recommend research-icp.npx claudepluginhub hungv47/meta-skills --plugin forsvn-devDesigns end-to-end user experiences: task flows, multi-step workflows, navigation, onboarding, signup, checkout, dashboards, error recovery, and cross-platform adaptation. Trigger when mapping user journeys or optimizing interactions.
Designs user flows and screen structures using StyleSeed UX patterns like progressive disclosure, hub-and-spoke navigation, and information pyramids. For onboarding, checkout, dashboards, and multi-screen features.
Maps analyzed screenshots into coherent interaction flows, navigation logic, and user journeys for apps and sites, identifying entry/exit points, gaps, and dead ends.