# Specification: Phase 6 Review and Regression Verification

## Overview
Perform a comprehensive review and technical audit of the Phase 6 implementation. This track focuses on ensuring zero regressions via a full-suite batch test run and expanding test coverage for the four major "Advanced Context" features introduced in Phase 6.

## Functional Requirements
- **Regression Testing:** Execute the entire test suite (230+ tests) in batches of 4 files each to verify codebase stability.
- **Feature Coverage:** Implement new integration and unit tests for:
    - **AST Tree Masking:** Verify per-symbol masking logic against unconventional C++ samples from the `gencpp` project (components and dependencies).
    - **Text Slice Highlighting:** Verify "Fuzzy Anchor" resolution resilience when files are modified.
    - **Takes Snapshotting:** Verify that `FileItem` states (flags, masks, slices) are perfectly restored when branching/switching discussions.
    - **Batch Operations UX:** Verify Shift-Click selection and mass-assignment logic.
- **Technical Audit:** Review the Phase 6 commits to ensure adherence to Conductor workflow protocols (TDD, MMA delegation, and documentation sync).

## Non-Functional Requirements
- **Stability:** Ensure 0 regressions across the full suite.
- **Documentation:** Synchronize `guide_architecture.md` and `guide_tools.md` with the new context mechanics.
- **Reporting:** Generate a final Phase 6 Verification Report summarizing the findings.

## Acceptance Criteria
- [ ] Full suite batch run completes with 100% pass rate.
- [ ] New tests for AST Masking, Text Slices, Snapshots, and Batch Ops are implemented and passing.
- [ ] Verification of AST parsing successful against `gencpp` source samples.
- [ ] Documentation updated to reflect the "Advanced Context Curation" architecture.
- [ ] Final Audit Report created and approved.

## Out of Scope
- Functional changes to the RAG engine.
- Implementation of new features beyond the Phase 6 scope.
- Automated MMA orchestration for the `gencpp` project itself (this track is strictly for Manual Slop infrastructure verification).