borghei/Claude-Skills / project-management/discovery/ai-prototyping
ai-prototyping
UnverifiedIdea to AI-generated prototype to customer validation to engineering handoff. Use when deciding prototype vs spec, choosing fidelity, briefing AI prototyping tools, testing with users, or handing a prototype to engineering.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 16.2 KB |
| assets/handoff_complete.json | config | 3.5 KB |
| assets/handoff_incomplete.json | config | 867 B |
| assets/handoff_template.md | markdown | 1.7 KB |
| assets/hypotheses_sample.json | config | 2.0 KB |
| assets/hypotheses_untestable.json | config | 1.2 KB |
| assets/prototype_brief_template.md | markdown | 2.7 KB |
| examples/mobile-expense-approvals.md | markdown | 6.1 KB |
| references/prototype-decision-guide.md | markdown | 8.0 KB |
| references/testing-handoff-governance.md | markdown | 8.9 KB |
| scripts/prototype_handoff_checker.py | script | 9.5 KB |
| scripts/prototype_plan.py | script | 10.2 KB |
assets/handoff_complete.json
View on GitHub{
"initiative": "Expense approvals on mobile",
"as_of": "2026-09-21",
"prototype_link": "https://example.internal/prototypes/mobile-approvals-v3",
"disposition": "throwaway - engineering rebuilds on the production mobile stack; reuse copy and flow only",
"problem": "Finance approvers at 50-500 employee customers batch approvals into a weekly desktop session, so employees wait a median of 6 days for reimbursement and submit duplicate chasers to support.",
"target_users": "Finance approvers (primary), submitting employees (secondary)",
"success_metrics": [
{"name": "Median submit-to-approval time", "baseline": "6 days", "target": "2 days", "window": "90 days after GA"},
{"name": "Share of approvals done on mobile", "baseline": "0%", "target": "30%", "window": "90 days after GA"}
],
"validated_findings": [
{"hypothesis": "H1 desirability", "participants": 10, "threshold": "at least 6 of 10 join pilot waitlist", "result": "7 of 10 joined", "status": "validated"},
{"hypothesis": "H2 usability", "participants": 6, "threshold": "at least 5 of 6 unassisted, median under 60s", "result": "5 of 6 unassisted, median 48s; 1 failure on the policy-violation badge", "status": "validated with fix"},
{"hypothesis": "H3 feasibility", "participants": 0, "threshold": "at least 90 percent precision", "result": "not tested in prototype - carried as a risk", "status": "open"}
],
"throwaway_list": [
"Auth: prototype uses a hard-coded demo login - build on the production SSO/session layer",
"Error states: offline, timeout and partial-sync states are not designed",
"Accessibility: generated UI not audited; screen reader labels missing on the receipt viewer",
"Security: no input validation, no rate limiting, no security review of generated code",
"Performance: synthetic data set of 40 receipts; no pagination or image compression",
"Data model: flat JSON mock - engineering designs the real schema and audit trail"
],
"acceptance_criteria": [
"Given a pending expense, when the approver taps the push notification, then the expense detail opens in under 2 seconds on a mid-range device",
"Given an out-of-policy line, when the approver views the expense, then the violation badge and the policy rule text are shown before the approve button",
"Given the approver rejects, when they submit, then a reason is mandatory and the employee is notified within 1 minute",
"All approval screens must meet WCAG 2.2 AA for contrast, labels and keyboard/switch access"
],
"non_goals": [
"Submitting expenses on mobile (already shipped)",
"Multi-level approval chains (desktop only in v1)",
"Offline approvals"
],
"constraints": ["Ship on the existing mobile app release train", "No new vendor for OCR in v1"],
"risks": ["H3 OCR precision unproven on real receipts - spike before sprint 2", "Push notification opt-in rates on Android unknown"],
"data_privacy": {
"real_customer_data_used": false,
"approval_ref": "",
"notes": "Prototype used 40 synthetic receipts generated from public templates; no customer receipts were uploaded to any prototype tool"
},
"security": "Prototype repository archived read-only; no production secrets were used; generated code will not be merged - rebuild follows the standard secure-SDLC review",
"accessibility": "Baseline WCAG 2.2 AA for the production build; prototype findings list known gaps above",
"ip_review": "Prototype tool terms reviewed by legal on 2026-09-02: outputs owned by us; no third-party assets or fonts with restrictive licenses in the prototype"
}
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the work | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixes | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation code | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |