Install
$ agentstack add skill-omexit-claude-skills-pack-user-flow ✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
About
User flow
What I'll do
Map the complete user journey for a feature — from entry point through completion — including happy paths, error states, edge cases, and decision points.
> user-flow vs flow-map: This skill maps the user's perspective — what users see, decide, and experience. > Use /flow-map for the system's perspective — internal states, network failures, concurrency, and cleanup paths. > Typical order: /user-flow first (understand the UX), then /flow-map (map all system paths for implementation).
Inputs I'll use (ask only if missing)
- The user goal (what are they trying to accomplish?)
- Entry points (how do they get here?)
- User persona (new vs returning, role, permissions)
- Constraints (platforms, accessibility requirements)
How I'll think about this
- Start with the user's goal, not the UI: "User wants to invite a teammate" not "User clicks the invite button." The goal determines the flow; the UI implements it.
- Map every state, not just the happy path: Loading, empty, partial, error, success, and "nothing happened" states. Users spend more time in non-happy states than you think.
- Identify decision points: Where does the user make a choice? What information do they need to make it? What happens if they choose wrong? Can they undo it?
- Trace error recovery: When something fails, can the user understand what went wrong and fix it without starting over? Error states without recovery paths are dead ends.
- Consider entry and exit points: Users don't always start at the beginning. They arrive via deep links, notifications, emails, search results. They leave mid-flow. Design for interruption and resumption.
- Check permission boundaries: What happens when a user without the right permissions reaches a step? Do they see a helpful message or a broken page?
Anti-patterns to flag
- Flows that only consider the happy path
- Error states that say "Something went wrong" without actionable guidance
- Flows that can't be resumed after interruption (lost form data, expired state)
- Decision points without enough context for the user to choose
- Accessibility dead ends (keyboard traps, missing screen reader context)
- Assuming all users enter from the home page
Quality bar
- Every state is accounted for: loading, empty, partial, error, success
- Error states include recovery paths (not just error messages)
- Decision points list what information the user needs
- Edge cases are identified (permissions, concurrent edits, expired sessions, slow connections)
- Accessibility considerations are noted at each step
- Flow can be handed to a designer or engineer without ambiguity
Workflow context
- Typically follows:
/prd(UX requirements) - Feeds into:
/design-doc(technical implementation),/test-plan(test scenarios),/ux-review - Related:
/experiment-design(measuring flow effectiveness)
Output
Fill templates/user-flow.md.
Learning & Memory
After completing this skill, store reusable insights in memory:
- User journey patterns: Common navigation flows, entry/exit points, and multi-step interaction sequences that recur across features
- Common friction points: Recurring UX obstacles such as confusing decision points, missing context, or broken recovery paths
- Error state patterns: Effective error recovery designs, dead-end patterns to avoid, and permission boundary handling strategies
Output contract
produces:
- type: "user-flow"
format: "markdown"
path: "claudedocs/-user-flow.md"
sections: [journeys, decision_points, error_states, edge_cases]
handoff: "Write claudedocs/handoff-user-flow-.yaml — suggest: flow-map, ux-review, ui-design"
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: OmexIT
- Source: OmexIT/claude-skills-pack
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet — be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.