Petal Coach

Situation

While reading What Color Is Your Parachute? for my job search, I found the Flower Exercise (a structured self-inventory across seven career dimensions) powerful but hard to work through on paper. The process was scattered, progress was easy to lose, and the real insights (recurring patterns across dimensions) were difficult to see without stepping back and synthesizing everything manually.

Task

Design and build a guided web app that walks users through all seven petals, persists progress locally, and optionally uses AI to surface patterns and generate coaching reflections, while keeping the core exercise fully functional without AI.

Action

• Wrote a full feature specification with 8 user stories, 28 functional requirements, and acceptance criteria using GitHub's spec-kit methodology
• Ran 3 rounds of clarification to resolve ambiguities (within-petal navigation, export eligibility thresholds, AI reflection triggers, multi-tab behavior)
• Completed a 57-item pre-implementation quality checklist catching gaps in edge cases, accessibility, and cross-requirement consistency
• Made deliberate architectural decisions: IndexedDB for client-side data ownership (no backend DB), server-side-only AI calls (keys never touch the client), provider-agnostic AI (Anthropic or OpenAI via env var)
• Implemented 46 tasks across 11 phases following the dependency-ordered task plan, validating at each checkpoint
• Built the skills petal with a story-based discovery flow: users write accomplishment stories, tag skills from the book's taxonomy, and the app computes recurring skills across stories

Result

Shipped a complete, production-ready app with all 8 user stories functional. Core exercise works with zero configuration. AI features (skill analysis, per-petal reflections, narrative synthesis, career brainstorming) activate automatically when API keys are set and degrade gracefully when not. Full PDF and Markdown export. The entire project, from spec to deployed code, demonstrates end-to-end product thinking: problem definition, requirements rigor, technical trade-offs, and execution.