wip: standalone code review feature and statusline API fix

- Add runStandaloneValidation() for validation outside task workflow
- Extract formatValidationResults() as shared function
- Refactor prepare-completion to use shared formatting
- Add code-review command and script (new feature)
- Fix statusline token counting for current_usage API change
- Add private journal entries

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

Author: cahaseler

.cc-track/backlog.md

@@ -13,3 +13,4 @@
 ## Items
 
 <!-- Items will be added below -->
+- [2025-12-14] Add new subagent to review agents tasked with identifying 'fallback' approaches, 'graceful degradation', and 'backwards compatibility' that simply masks failures with fake functionality or broken systems. Any backwards compatibility or fallbacks MUST be in the spec. Falling back to something that appears at first glance to be working but isn't makes it substantially harder to test AI generated code.

.cc-track/reviews/2025-12-14.md

@@ -0,0 +1,52 @@
+# Code Review: 2025-12-14
+
+**Branch:** main
+**Generated:** 2025-12-14
+**Status:** resolved
+
+## Summary
+
+| # | Score | Status | Location | Decision | Action |
+|---|-------|--------|----------|----------|--------|
+| 1 | 100 | resolved | code-review.ts:89-98 | Fix | Changed to success:false, exitCode:1 |
+| 2 | 75 | resolved | validation.ts | Fix | Extracted runCoreValidationChecks() |
+| 3 | 75 | resolved | code-review.test.ts | Fix | Added 15 unit tests |
+| 4 | 75 | resolved | validation.ts | Fix | Extracted formatValidationResults() |
+
+## Issues
+
+### Issue 1: Error handler returns success:true with exitCode:0
+- **Score:** 100
+- **Location:** code-review.ts:89-98
+- **Reported by:** bug-scanner, guidelines-reviewer
+- **Description:** The catch block returns success:true with exitCode:0 when validation fails, violating the CommandFailure interface contract and causing CI pipelines to misinterpret errors as success.
+- **Status:** resolved
+- **Decision:** Fix
+- **Action:** Changed catch block to return success:false and exitCode:1
+
+### Issue 2: Code duplication in validation functions
+- **Score:** 75
+- **Location:** validation.ts
+- **Reported by:** duplication-detector
+- **Description:** runValidationChecks() and runStandaloneValidation() have 95%+ identical code (60+ lines). Both execute same validation logic with only minor differences.
+- **Status:** resolved
+- **Decision:** Fix
+- **Action:** Extracted runCoreValidationChecks() helper that both functions now use
+
+### Issue 3: Missing unit tests for code-review.ts
+- **Score:** 75
+- **Location:** code-review.test.ts
+- **Reported by:** guidelines-reviewer
+- **Description:** Constitution requires 80%+ coverage with colocated test files. The 195-line script has no tests.
+- **Status:** resolved
+- **Decision:** Fix
+- **Action:** Created code-review.test.ts with 15 tests covering happy path, validation failures, error handling, and edge cases
+
+### Issue 4: Presentation logic duplication
+- **Score:** 75
+- **Location:** validation.ts
+- **Reported by:** duplication-detector
+- **Description:** ~70 lines of validation result formatting logic duplicated across both scripts.
+- **Status:** resolved
+- **Decision:** Fix
+- **Action:** Extracted formatValidationResults() to validation.ts, updated both code-review.ts and prepare-completion.ts to use it

.claude/settings.json

@@ -5,7 +5,7 @@
   },
   "statusLine": {
     "type": "command",
-    "command": "bun run ${CLAUDE_PLUGIN_ROOT}/scripts/statusline.ts",
+    "command": "bun run /home/ubuntu/projects/cc-pars/scripts/statusline.ts",
     "padding": 0
   }
 }