chore: Project cleanup and Canvas UX improvements (Phase 7-9)
## Cleanup (v0.5.0) - Delete 102+ orphaned MCP session temp files - Remove build artifacts (htmlcov, dist, __pycache__) - Archive superseded plan docs (RALPH_LOOP V2/V3, CANVAS V3, etc.) - Move debug/analysis scripts from tests/ to tools/analysis/ - Archive redundant NX journals to archive/nx_journals/ - Archive monolithic PROTOCOL.md to docs/archive/ - Update .gitignore with missing patterns - Clean old study files (optimization_log_old.txt, run_optimization_old.py) ## Canvas UX (Phases 7-9) - Phase 7: Resizable panels with localStorage persistence - Left sidebar: 200-400px, Right panel: 280-600px - New useResizablePanel hook and ResizeHandle component - Phase 8: Enable all palette items - All 8 node types now draggable - Singleton logic for model/solver/algorithm/surrogate - Phase 9: Solver configuration - Add SolverEngine type (nxnastran, mscnastran, python, etc.) - Add NastranSolutionType (SOL101-SOL200) - Engine/solution dropdowns in config panel - Python script path support ## Documentation - Update CHANGELOG.md with recent versions - Update docs/00_INDEX.md - Create examples/README.md - Add docs/plans/CANVAS_UX_IMPROVEMENTS.md
This commit is contained in:
43
CLAUDE.md
43
CLAUDE.md
@@ -55,6 +55,49 @@ If working directory is inside a study (`studies/*/`):
|
||||
- If no study context: Offer to create one or list available studies
|
||||
- After code changes: Update documentation proactively (SYS_12, cheatsheet)
|
||||
|
||||
### Step 5: Use DevLoop for Multi-Step Development Tasks
|
||||
|
||||
**CRITICAL: For any development task with 3+ steps, USE DEVLOOP instead of manual work.**
|
||||
|
||||
DevLoop is the closed-loop development system that coordinates AI agents for autonomous development:
|
||||
|
||||
```bash
|
||||
# Plan a task with Gemini
|
||||
python tools/devloop_cli.py plan "fix extractor exports"
|
||||
|
||||
# Implement with Claude
|
||||
python tools/devloop_cli.py implement
|
||||
|
||||
# Test filesystem/API
|
||||
python tools/devloop_cli.py test --study support_arm
|
||||
|
||||
# Test dashboard UI with Playwright
|
||||
python tools/devloop_cli.py browser --level full
|
||||
|
||||
# Analyze failures
|
||||
python tools/devloop_cli.py analyze
|
||||
|
||||
# Full autonomous cycle
|
||||
python tools/devloop_cli.py start "add new stress extractor"
|
||||
```
|
||||
|
||||
**When to use DevLoop:**
|
||||
- Fixing bugs that require multiple file changes
|
||||
- Adding new features or extractors
|
||||
- Debugging optimization failures
|
||||
- Testing dashboard UI changes
|
||||
- Any task that would take 3+ manual steps
|
||||
|
||||
**Browser test levels:**
|
||||
- `quick` - Smoke test (1 test)
|
||||
- `home` - Home page verification (2 tests)
|
||||
- `full` - All UI tests (5+ tests)
|
||||
- `study` - Canvas/dashboard for specific study
|
||||
|
||||
**DO NOT default to manual debugging** - use the automation we built!
|
||||
|
||||
**Full documentation**: `docs/guides/DEVLOOP.md`
|
||||
|
||||
---
|
||||
|
||||
## Quick Start - Protocol Operating System
|
||||
|
||||
Reference in New Issue
Block a user