All checks were successful
Deploy to Production / Build and Deploy (push) Successful in 12s
- Sidebar: dynamic brand-accent colors, brainstorm section restyled - AI chat general: popup panel with expand/collapse, hides when contextual AI open - AI chat contextual: tabs reordered (Actions first), X close button, height fix - Settings: all tabs restyled, 6 new color presets (sage, terracotta, iron, etc.) - Global color cleanup: emerald/orange hardcoded → brand-accent dynamic - Brainstorm page: orange → brand-accent throughout - PageEntry animation component added to key pages - Floating AI button: bg-brand-accent instead of hardcoded black - i18n: all 15 locales updated with new AI/billing keys - Billing: freemium quota tracking, BYOK, stripe subscription scaffolding - Admin: integrated into new design - AGENTS.md + CLAUDE.md project rules added
1.4 KiB
1.4 KiB
Session Start Protocol
When starting or resuming a session, always follow this sequence before implementing anything:
1. Read the Dialog Document
Read the dialog file completely to understand:
- What steps are done
- What steps remain
- Any blockers or change requests
- Current context and decisions
2. Verify Plan Against Reality
The plan may be outdated. Check if:
- Steps marked "To Do" have actually been implemented
- Steps marked "Done" are truly complete
- Numbering is sequential and accurate
If the plan is outdated → Update it before proceeding.
3. Present Current Status
Summarize for the designer:
- What's done (with step numbers)
- What's remaining (with step numbers)
- Any change requests pending
4. Before Implementing a Step
Always check the specification/sketches first:
Agent: "Before implementing step 20, let me check the sketches..."
Agent: "I see this requires a nested drawer pattern, not inline buttons.
Should I break this into sub-steps?"
This prevents building the wrong thing and wasting effort.
Why This Matters
Sessions can be interrupted. Context can be lost. The dialog document survives — but only if it's kept accurate. This protocol ensures:
- No duplicate work (re-implementing what exists)
- No missed work (skipping what's actually needed)
- Correct understanding of requirements before implementation