NullPointerException

Mise en Place is a cooking co-pilot that turns recipes into guided, voice-driven sessions. It ingests recipe data, lets users pick a dish, then walks them through prep and cooking in a structured “cooking mode” with timers, step progression, and voice prompts powered by an AI agent. The submission shows the full flow—from recipe list to stepwise guidance—demonstrating an AI-driven assistant embedded in a Flutter app.

How to run the demo end-to-end
Prereqs: Flutter/Dart SDK (respecting .fvmrc if you use FVM), a mobile emulator or device.
Install deps: flutter pub get in /mise-en-place/prototype.
Run: flutter run (select simulator/device).
For voice agent (if demonstrating ElevenLabs integration): configure elevenlabs/agent_config.json with your API keys and run the provided push_agent.sh/pull_agent.sh scripts to sync the agent before launching the app.
Demo path: Open the app → recipe list → select a recipe → start cooking mode → experience step-by-step guidance with timers and voice prompts.

Judging criteria fit
Working Prototype: End-to-end Flutter app with functioning recipe list, detail view, and interactive cooking mode; timers and step navigation are live, not mocked.
Technical Complexity & Integration: Combines Flutter UI/state management with an AI voice agent (ElevenLabs) via configurable agent_config.json; shell scripts for agent deploy/sync; structured controller logic (cooking_mode_controller.dart) orchestrates timers, steps, and prompts.
Innovation & Creativity: Transforms static recipes into an interactive, voice-assisted workflow; blends real-time guidance, pacing, and safety reminders rather than just displaying text.
Real-World Impact: Reduces kitchen friction and cognitive load, supports hands-free cooking, and helps novices complete recipes correctly with timely cues.
Theme Alignment: Demonstrates an AI “agent” embedded in a consumer flow (cooking), showcasing applied AI assistance and multimodal interaction.

Notable technologies and services
Flutter/Dart (mobile app UI and state)
ElevenLabs voice agent (speech synthesis/agent orchestration via agent_config.json)
Shell utilities (push_agent.sh, pull_agent.sh) to manage agent configs
(If applicable) Any recipe data source or local JSON/Firestore—note whichever you used

Reproducibility checklist
Provide API keys/env vars for ElevenLabs in agent_config.json (redact in repo; share with judges separately).
Confirm FVM version from .fvmrc or use system Flutter that matches your tested version.
Ensure device/emulator has audio output to hear prompts.
Optional: include a short screen recording of the recipe → cooking mode flow and a sample agent response.