docs: update meta files for conflict detection session

TODO.md

@@ -12,9 +12,12 @@
 - [ ] **Agent Teams real-world testing** -- Env var whitelist fix done. 3 test sessions ran ($1.10, $0.69, $1.70) but model didn't spawn subagents — needs complex multi-part prompts to trigger delegation. Test with CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1.
 - [ ] **Configurable stall threshold** -- health.yaml per-project config for stall threshold (currently hardcoded 15 min). Adaptive suggestions after 50 sessions from session_metrics data.
 - [ ] **Agent provider adapter pattern** -- Abstract model interface (ProviderConfig, ProviderCapabilities) for multi-model support (Claude, Codex, Ollama). Rename agent-runner.mjs → claude-runner.mjs, capability gates in UI.
+- [ ] **S-1 Phase 2 — filesystem write detection** -- inotify-based file change detection (already have notify crate). PID attribution to distinguish BTerminal-managed vs external writes. Always-on red "write conflict" cards.
+- [ ] **S-2 — session anchors** -- Preserve first 5 turns / 2K tokens verbatim through checkpoint chains. Pin additional turns via AST conversation tree action.
 
 ## Completed
 
+- [x] **File Overlap Conflict Detection (S-1 Phase 1)** -- conflicts.svelte.ts store tracks per-session Write/Edit file paths. Detects 2+ agents writing same file. Health attention score=70. Toast on conflict. Red badge in ProjectHeader + StatusBar. Shared tool-files.ts utility. 170/170 tests. | Done: 2026-03-10
 - [x] **Project Health Dashboard (S-3)** -- health.svelte.ts store (activity state, burn rate, context pressure, attention scoring). Mission Control status bar (running/idle/stalled counts, $/hr, attention queue). ProjectHeader indicators (status dot, ctx%, $/hr). session_metrics SQLite table (100-row retention). Metric persistence on agent completion. | Done: 2026-03-10
 - [x] **Context tab repurpose** -- Replaced ContextPane (ctx database viewer) with ContextTab (LLM context window visualization). Stats bar, segmented token meter, file references, turn breakdown. Tribunal debate (S-1-R4 at 82%). Token estimation via ~4 chars/token heuristic. | Done: 2026-03-10
 - [x] **CodeMirror 6 editor** -- Replaced shiki viewer in FilesTab with CodeMirror 6. CodeEditor.svelte wrapper: 15 lazy-loaded language modes, Catppuccin theme, auto-close brackets, bracket matching, code folding, Ctrl+S save, dirty tracking, save-on-blur setting. write_file_content Rust command. | Done: 2026-03-10
@@ -23,5 +26,4 @@
 - [x] **AgentPane collapsibles + aspect ratio** -- Text messages collapsible (open by default), cost summary collapsed by default. Project max aspect ratio setting (0.3–3.0, default 1.0) with CSS var + SettingsTab stepper. No-implicit-push rule 52. Desktop StartupWMClass. | Done: 2026-03-09
 - [x] **AgentPane + MarkdownPane UI redesign** -- AgentPane: sans-serif font, tool call/result pairing via $derived.by, hook message collapsing, context meter, muted colors via color-mix(), responsive margins via container queries. MarkdownPane: container query wrapper, shared responsive padding variable. Tribunal-elected design (S-3-R4, 88% confidence). 139/139 tests pass. | Done: 2026-03-09
 - [x] **E2E testing — consolidated & expanded** -- Consolidated 4 spec files into single bterminal.test.ts (Tauri single-session requirement). 25 tests across 4 describe blocks: Smoke(6), Workspace(8), Settings(6), Keyboard(5). Fixed WebDriver clicks on Svelte components via browser.execute(), removed tauri-plugin-log (redundant with telemetry::init()). | Done: 2026-03-08
-- [x] **px→rem conversion** -- All ~100 px layout violations converted to rem across 10 components (AgentPane, ToastContainer, CommandPalette, SettingsTab, TeamAgentsPanel, AgentCard, StatusBar, AgentTree, TerminalPane, AgentPreviewPane). Rule 18 fully enforced. | Done: 2026-03-08
 - [x] **Workspace teardown race fix** -- Added pendingPersistCount counter + waitForPendingPersistence() fence in agent-dispatcher.ts. switchGroup() awaits persistence before clearing state. Last open HIGH audit finding resolved. | Done: 2026-03-08