Merge branch 'main' into claude/plan-desktop-app-aaSsY

Author: bkauto3

.github/workflows/python-verify.yml

@@ -0,0 +1,25 @@
+name: Python Verify
+
+on:
+  push:
+    branches:
+      - main
+  pull_request:
+
+jobs:
+  verify:
+    runs-on: windows-latest
+    steps:
+      - name: Check out repository
+        uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+
+      - name: Install package and test dependencies
+        run: python -m pip install --upgrade pip && python -m pip install .[dev]
+
+      - name: Verify Python source and built wheel
+        run: python scripts/verify_python_build.py

.gitignore

@@ -40,8 +40,50 @@ CATO_SESSION_*.md
 *_SESSION_*.md
 *_EXECUTION.md
 *_IMPLEMENTATION_REPORT.md
+
+# Dev session docs and checklists (personal working files)
+CATO_BUILD_IMPROVEMENTS_CHECKLIST.md
+CATO_BUILD_POTENTIAL_ISSUES.md
+CATO_BUILD_RISKS.md
+CATO_CLI_INTEGRATION_CHECKLIST.md
+CATO_FINAL_TEST_REPORT.md
+CATO_GENESIS_BRAINSTORM.md
+CATO_HUDSON_REVIEW_M3.md
+CATO_HUDSON_REVIEW_M4.md
+DAEMON_RESTART_COMPLETE.md
+SKILLS_RECOGNITION_FIX_VERIFIED.md
+TELEGRAM_DESKTOP_INTEGRATION_FIX.md
+TELEGRAM_INTEGRATION_COMPLETE.md
+desktop/WINDOWS_BUILD.md
+
+# Kraken/test run artifacts
+kraken_e2e_report.json
+kraken_e2e_test.py
+kraken_test_output.txt
+kraken_test_run.txt
+test_skills.py
+
+# Local bridge/startup scripts (machine-specific)
+launch_bridge.py
+register_telegram_bridge_task.ps1
+start_telegram_bridge.ps1
+scripts/install_watchdog.ps1
+
+# Utility scripts (local dev only)
+scripts/sync_version.py
+scripts/verify_python_build.py
+
+# Claude Code session data
+.claude/
+
+# Log directories
+logs/
 # Desktop app build artifacts
 desktop/src-tauri/target/
 desktop/dist/
 desktop/node_modules/
 desktop/src-tauri/binaries/
+
+# Ralph Loop state directories
+Ralph Loop Desktop APP/
+Ralph Loop Web UI/

.ralph/guardrails.md

@@ -0,0 +1,33 @@
+# RALPH LOOP GUARDRAILS
+# Cato Integration Project
+
+## HARD RULES
+1. Tests MUST stay 100% passing after every iteration — check with: python -m pytest tests/ -x -q --tb=short
+2. Never rename existing Python modules (breaks imports + tests)
+3. Never modify existing working API routes — only ADD new ones
+4. Never break existing desktop views — only ADD/MODIFY what's needed
+5. Max 5 iterations per chunk before Hudson escalation
+6. All file writes use dedicated Write/Edit tools — never bash echo/heredoc
+7. All new TSX files follow React functional component pattern with named + default export
+8. Never introduce hardcoded IP/port — use httpPort prop
+9. All new server.py route handlers have try/except with logger.error + graceful fallback
+10. Input validation on all new routes that take path params (regex allowlist on names)
+
+## ITERATION PROTOCOL
+- Each iteration: implement → run tests → check output → fix if needed
+- If tests fail after 5 iterations: call Hudson agent, pass full error context
+- Hudson fixes, returns control, continue to next chunk
+
+## NAMING CONVENTIONS
+- Views: PascalCase + "View" suffix (MemoryView, SystemView, etc.)
+- Routes: lowercase hyphenated (/api/memory/stats, /api/cli/status)
+- State variables: camelCase
+- CSS classes: kebab-case matching existing patterns
+
+## DO NOT TOUCH
+- cato/orchestrator/clawflows.py (working, tests cover it)
+- cato/audit/* (working, tested)
+- cato/core/memory.py (working, tested)
+- Any file in tests/ (unless adding new tests)
+- pyproject.toml / Cargo.toml
+- conftest.py

.ralph/implementation_plan.md

@@ -0,0 +1,134 @@
+# CATO INTEGRATION IMPLEMENTATION PLAN
+# Generated: 2026-03-08
+# Owner: Ralph Wiggum Loop
+
+## OBJECTIVE
+Integrate all backend modules from cato/ into the Cato Desktop App and Web UI.
+Fix broken save buttons. Add missing views for every exposed API route.
+Max 5 iterations per chunk. Escalate to Hudson if chunk fails after 5 iterations.
+Backend-architect handles migration/wiring after each chunk. Alex assembles. Kraken audits live.
+
+---
+
+## CHUNKS
+
+### CHUNK 1 — Fix Broken Save Buttons (CRITICAL)
+**Files:** SkillsView.tsx, IdentityView.tsx (investigate why save fails)
+**Tasks:**
+- [ ] Add Save button + save() fn to SkillsView.tsx (PATCH /api/skills/{name}/content)
+- [ ] Add "saved" toast/banner to SkillsView.tsx
+- [ ] Debug IdentityView.tsx save (button disabled when isDirty=false — investigate why)
+- [ ] Verify workspace_put route in server.py returns correct JSON
+- [ ] Test both saves end-to-end against live daemon
+**Escalation:** Hudson if >5 iterations
+
+### CHUNK 2 — Memory View
+**Files:** NEW desktop/src/views/MemoryView.tsx, server.py (routes already exist)
+**Tasks:**
+- [ ] Create MemoryView.tsx with tabs: Facts | Chunks | Stats | Graph
+- [ ] Facts tab: list stored facts, search, delete individual facts
+- [ ] Stats tab: total facts, kg_nodes, kg_edges, embedding model, DB size
+- [ ] Chunks tab: recent context chunks with token counts
+- [ ] Wire /api/memory/files, /api/memory/content, /api/memory/stats
+- [ ] Add "Memory" to Sidebar nav (Monitoring group)
+- [ ] Update App.tsx with MemoryView import + case
+**Escalation:** Hudson if >5 iterations
+
+### CHUNK 3 — CLI Process Pool Monitor + Action Guard Dashboard
+**Files:** NEW desktop/src/views/SystemView.tsx
+**Tasks:**
+- [ ] Create SystemView.tsx with two panels:
+  - Panel A: CLI Process Pool (GET /api/cli/status) — warm/cold status, model PIDs
+  - Panel B: Action Guard (GET /api/action-guard/status) — 3-rule gate status, autonomy level
+  - Panel C: Daemon restart button (POST /api/daemon/restart)
+- [ ] Add "System" to Sidebar nav (Settings group)
+- [ ] Update App.tsx
+**Escalation:** Hudson if >5 iterations
+
+### CHUNK 4 — Replay Visualization
+**Files:** NEW desktop/src/views/ReplayView.tsx, update SessionsView.tsx
+**Tasks:**
+- [ ] Create ReplayView.tsx — shows step-by-step replay with match/mismatch
+- [ ] Add "Replay" button to SessionsView.tsx that triggers POST /api/sessions/{id}/replay
+- [ ] Navigate to ReplayView with results after replay completes
+- [ ] Show: total steps, matched %, mismatched list, timing per step
+- [ ] Update App.tsx + Sidebar
+**Escalation:** Hudson if >5 iterations
+
+### CHUNK 5 — Skill Improvement + Diagnostics Panel
+**Files:** server.py (new routes), NEW desktop/src/views/DiagnosticsView.tsx
+**Tasks:**
+- [ ] Add GET /api/diagnostics/query-classifier — expose QueryClassifier tier decisions
+- [ ] Add GET /api/diagnostics/skill-corrections — list stored corrections
+- [ ] Add GET /api/diagnostics/contradiction-health — ContradictionDetector health summary
+- [ ] Add GET /api/diagnostics/decision-memory — DecisionMemory open decisions + overconfidence
+- [ ] Add GET /api/diagnostics/anomaly-domains — AnomalyDetector domain summaries
+- [ ] Create DiagnosticsView.tsx with tabs for each diagnostic
+- [ ] Add "Diagnostics" to Sidebar nav (Monitoring group)
+- [ ] Update App.tsx
+**Escalation:** Hudson if >5 iterations
+
+### CHUNK 6 — Adapter Status + Heartbeat View
+**Files:** server.py (new routes), update DashboardView.tsx
+**Tasks:**
+- [ ] Add GET /api/adapters — list active adapters (Telegram, WhatsApp) and their status
+- [ ] Add GET /api/heartbeat — last heartbeat timestamp + agent name + uptime
+- [ ] Add adapter status pills to DashboardView.tsx
+- [ ] Add heartbeat info to DashboardView.tsx
+- [ ] Wire Telegram adapter status to ConfigView.tsx (already has Telegram section)
+**Escalation:** Hudson if >5 iterations
+
+### CHUNK 7 — Session Checkpoints + Receipts
+**Files:** server.py (new routes), update SessionsView.tsx, AuditLogView.tsx
+**Tasks:**
+- [ ] Add GET /api/sessions/{id}/checkpoints — list checkpoints for session
+- [ ] Add GET /api/sessions/{id}/checkpoints/{checkpoint_id} — get checkpoint summary
+- [ ] Add GET /api/sessions/{id}/receipt — generate signed receipt (receipt.py)
+- [ ] Add "Checkpoints" tab to SessionsView.tsx
+- [ ] Add "Download Receipt" button to AuditLogView.tsx
+- [ ] Update SessionsView.tsx to show checkpoint count per session
+**Escalation:** Hudson if >5 iterations
+
+### CHUNK 8 — Web UI (dashboard.html + coding_agent.html) Sync
+**Files:** cato/ui/dashboard.html, cato/ui/coding_agent.html
+**Tasks:**
+- [ ] Audit dashboard.html for all nav items — add Memory, System, Diagnostics, Replay pages
+- [ ] Add JS handlers for Memory page (facts table, stats, search)
+- [ ] Add JS handlers for System page (CLI pool status, action guard, daemon restart)
+- [ ] Add JS handlers for Diagnostics page
+- [ ] Fix MODEL_ICONS in coding_agent.html (C/X/G/⊕) — keep as text abbreviations (these are intentional)
+- [ ] Verify coding_agent.html has logo in header (currently favicon only)
+**Escalation:** Hudson if >5 iterations
+
+---
+
+## INTEGRATION RULES
+
+1. All new views follow existing patterns (fetch on mount, auto-refresh where needed)
+2. All new API routes follow existing handler pattern in server.py
+3. All new views exported as named + default export
+4. All new routes registered in router section of server.py
+5. Tests must stay 100% passing after every chunk
+6. No hardcoded ports — use httpPort prop
+7. Error states shown with existing .error-banner / .page-error classes
+8. Loading states shown with .app-loading-spinner
+
+---
+
+## AGENT HANDOFF PROTOCOL
+
+```
+Ralph Loop → chunk complete
+     ↓
+backend-architect → migration/wiring verification
+     ↓
+Next chunk (Ralph Loop)
+     ↓
+[After all chunks]
+     ↓
+Alex → assemble + integration check
+     ↓
+Kraken → live audit + fix
+     ↓
+DONE
+```