taskplane 0.28.3 → 0.28.5

package/dashboard/server.cjs

@@ -1,1666 +1,1666 @@
-#!/usr/bin/env node
-/**
- * Orchestrator Web Dashboard — Local HTTP server with SSE live updates.
- *
- * Reads .pi/batch-state.json + STATUS.md files and streams state to the
- * browser via Server-Sent Events. Zero external dependencies.
- *
- * Usage:
- *   node dashboard/server.cjs [--port 8099] [--root /path/to/project]
- */
-
-const http = require("http");
-const fs = require("fs");
-const path = require("path");
-const { exec } = require("child_process");
-// url module not needed — we parse with new URL() below
-
-// ─── Configuration ──────────────────────────────────────────────────────────
-
-const PUBLIC_DIR = path.join(__dirname, "public");
-const DEFAULT_PORT = 8100;
-const MAX_PORT_ATTEMPTS = 20;
-const POLL_INTERVAL = 2000; // ms between state checks
-
-// REPO_ROOT is resolved after parseArgs() — see initialization below.
-// In workspace mode, REPO_ROOT is the workspace root (passed via --root).
-// All dashboard state paths (batch-state, lane-state, conversation logs,
-// batch-history) live at <REPO_ROOT>/.pi/ — this is runtime/sidecar state
-// which does NOT follow the taskplane-pointer.json resolution chain.
-// The pointer directs config/agent lookups to a config repo, but the
-// dashboard only reads state files, so no pointer resolution is needed here.
-let REPO_ROOT;
-let BATCH_STATE_PATH;
-let BATCH_HISTORY_PATH;
-
-// ─── CLI Args ───────────────────────────────────────────────────────────────
-
-function parseArgs() {
-  const args = process.argv.slice(2);
-  const opts = { port: DEFAULT_PORT, open: true, root: "" };
-  for (let i = 0; i < args.length; i++) {
-    if (args[i] === "--port" && args[i + 1]) {
-      opts.port = parseInt(args[i + 1]) || DEFAULT_PORT;
-      i++;
-    } else if (args[i] === "--root" && args[i + 1]) {
-      opts.root = args[i + 1];
-      i++;
-    } else if (args[i] === "--no-open") {
-      opts.open = false;
-    } else if (args[i] === "--help" || args[i] === "-h") {
-      console.log(`
-Orchestrator Web Dashboard
-
-Usage:
-  node dashboard/server.cjs [options]
-
-Options:
-  --port <number>   Port to listen on (default: ${DEFAULT_PORT})
-  --root <path>     Project root directory (default: current directory)
-  --no-open         Don't auto-open browser
-  -h, --help        Show this help
-`);
-      process.exit(0);
-    }
-  }
-  return opts;
-}
-
-// ─── Data Loading (ported from orch-dashboard.cjs) ──────────────────────────
-
-function normalizeBatchStateIngress(state) {
-  if (!state || typeof state !== "object" || !Array.isArray(state.lanes)) {
-    return state;
-  }
-
-  for (const lane of state.lanes) {
-    if (!lane || typeof lane !== "object") continue;
-    // Legacy compatibility: older persisted states stored lane session IDs under
-    // `tmuxSessionName`. Normalize to `laneSessionId` at ingress and drop legacy key.
-    const laneSessionId = typeof lane.laneSessionId === "string"
-      ? lane.laneSessionId
-      : (typeof lane.tmuxSessionName === "string" ? lane.tmuxSessionName : undefined);
-    if (laneSessionId) {
-      lane.laneSessionId = laneSessionId;
-    }
-    if ("tmuxSessionName" in lane) {
-      delete lane.tmuxSessionName;
-    }
-  }
-
-  return state;
-}
-
-function loadBatchState() {
-  try {
-    const raw = fs.readFileSync(BATCH_STATE_PATH, "utf-8");
-    return normalizeBatchStateIngress(JSON.parse(raw));
-  } catch {
-    return null;
-  }
-}
-
-function resolveTaskFolder(task, state) {
-  if (!task || !task.taskFolder) return null;
-  const laneNum = task.laneNumber;
-  const lane = (state?.lanes || []).find((l) => l.laneNumber === laneNum);
-  if (!lane || !lane.worktreePath) return task.taskFolder;
-
-  // In workspace mode, the worktree is inside a specific repo, not the workspace root.
-  // The task folder path needs to be made relative to the repo root (parent of the worktree),
-  // not the workspace root. Detect this by finding the repo root from the worktree path.
-  const taskFolderAbs = path.resolve(task.taskFolder);
-  const worktreeAbs = path.resolve(lane.worktreePath);
-
-  // Try to find the repo root: walk up from the worktree path looking for which
-  // ancestor is a prefix of the task folder. The worktree is at <repoRoot>/.worktrees/<name>
-  // or a sibling, so the repo root is typically 2 levels up from a subdirectory worktree.
-  // Heuristic: find the longest common ancestor between taskFolder and worktree's repo root.
-  const repoRootAbs = path.resolve(REPO_ROOT);
-
-  // First try: relative to workspace root (works in repo mode where workspace = repo)
-  let rel = path.relative(repoRootAbs, taskFolderAbs);
-  if (!rel || rel.startsWith("..") || path.isAbsolute(rel)) return task.taskFolder;
-
-  // Check if joining with worktree produces a valid path
-  const candidate = path.join(worktreeAbs, rel);
-  try {
-    if (fs.existsSync(candidate)) return candidate;
-  } catch { /* fall through */ }
-
-  // Second try: the worktree is inside a repo subdirectory of the workspace root.
-  // Strip the repo prefix from the task folder path to get the repo-relative path.
-  // e.g., taskFolder = "workspace/platform-docs/task-mgmt/DOC-001/"
-  //        worktree  = "workspace/platform-docs/.worktrees/wt-1/"
-  //        repo-relative = "task-mgmt/DOC-001/"
-  // Find the repo by checking which workspace repo path is a prefix of the task folder.
-  const repoRoots = [];
-  try {
-    const stateMode = state.mode;
-    if (stateMode === "workspace" && state.repos) {
-      for (const r of state.repos) repoRoots.push(path.resolve(r.path));
-    }
-  } catch { /* no repo info in state */ }
-
-  // Also try inferring repo root from worktree path pattern:
-  // .worktrees/<name> → parent is repo root; sibling worktrees → shared parent
-  const worktreeParent = path.dirname(worktreeAbs);
-  const worktreeGrandparent = path.dirname(worktreeParent);
-  for (const possibleRepoRoot of [worktreeGrandparent, ...repoRoots]) {
-    const repoRel = path.relative(possibleRepoRoot, taskFolderAbs);
-    if (repoRel && !repoRel.startsWith("..") && !path.isAbsolute(repoRel)) {
-      const repoCandidate = path.join(worktreeAbs, repoRel);
-      try {
-        if (fs.existsSync(repoCandidate)) return repoCandidate;
-      } catch { continue; }
-    }
-  }
-
-  // Fallback: return original task folder (might work if not in worktree)
-  return task.taskFolder;
-}
-
-function parseStatusMd(taskFolder) {
-  const candidates = [taskFolder];
-  const taskId = path.basename(taskFolder);
-  const archiveBase = taskFolder.replace(/[/\\]tasks[/\\][^/\\]+$/, "/tasks/archive/" + taskId);
-  if (archiveBase !== taskFolder) candidates.push(archiveBase);
-
-  for (const folder of candidates) {
-    const statusPath = path.join(folder, "STATUS.md");
-    try {
-      const content = fs.readFileSync(statusPath, "utf-8");
-      const stepMatch = content.match(/\*\*Current Step:\*\*\s*(.+)/);
-      const statusMatch = content.match(/\*\*Status:\*\*\s*(.+)/);
-      const iterMatch = content.match(/\*\*Iteration:\*\*\s*(\d+)/);
-      const reviewMatch = content.match(/\*\*Review Counter:\*\*\s*(\d+)/);
-      const checked = (content.match(/- \[x\]/gi) || []).length;
-      const unchecked = (content.match(/- \[ \]/g) || []).length;
-      const total = checked + unchecked;
-      return {
-        currentStep: stepMatch ? stepMatch[1].trim() : "Unknown",
-        status: statusMatch ? statusMatch[1].trim() : "Unknown",
-        iteration: iterMatch ? parseInt(iterMatch[1]) : 0,
-        reviews: reviewMatch ? parseInt(reviewMatch[1]) : 0,
-        checked,
-        total,
-        progress: total > 0 ? Math.round((checked / total) * 100) : 0,
-      };
-    } catch {
-      continue;
-    }
-  }
-  return null;
-}
-
-function getActiveSessions() {
-  // Runtime V2: return active merger session names from the runtime registry
-  // so the dashboard merge pane can display live telemetry for running agents.
-  // Terminal statuses indicate the agent is no longer alive.
-  const TERMINAL_STATUSES = new Set(["exited", "killed", "crashed", "timed_out"]);
-  try {
-    const state = loadBatchState();
-    if (!state || !state.batchId) return [];
-    const registry = loadRuntimeRegistry(state.batchId);
-    if (!registry || !registry.agents) return [];
-    return Object.values(registry.agents)
-      .filter(a => a.role === "merger" && !TERMINAL_STATUSES.has(a.status))
-      .map(a => a.agentId);
-  } catch {
-    return [];
-  }
-}
-
-function checkDoneFile(taskFolder) {
-  const candidates = [taskFolder];
-  const taskId = path.basename(taskFolder);
-  const archiveBase = taskFolder.replace(/[/\\]tasks[/\\][^/\\]+$/, "/tasks/archive/" + taskId);
-  if (archiveBase !== taskFolder) candidates.push(archiveBase);
-  for (const folder of candidates) {
-    if (fs.existsSync(path.join(folder, ".DONE"))) return true;
-  }
-  return false;
-}
-
-/** Read lane state sidecar JSON files written by the task-runner. */
-function loadLaneStates() {
-  const piDir = path.join(REPO_ROOT, ".pi");
-  const states = {};
-  try {
-    const files = fs.readdirSync(piDir).filter(f => f.startsWith("lane-state-") && f.endsWith(".json"));
-    for (const file of files) {
-      try {
-        const raw = fs.readFileSync(path.join(piDir, file), "utf-8").trim();
-        if (!raw) continue;
-        const data = JSON.parse(raw);
-        if (data.prefix) states[data.prefix] = data;
-      } catch { continue; }
-    }
-  } catch { /* .pi dir may not exist */ }
-  return states;
-}
-
-// ─── Telemetry JSONL Tailing ────────────────────────────────────────────────
-
-/**
- * Module-level tail state for incremental JSONL reading.
- * Persists across poll ticks within this server process.
- * Key: absolute file path → { offset, partial }
- */
-const telemetryTailStates = new Map();
-
-/**
- * Module-level accumulated telemetry per session prefix.
- * Persists across poll ticks so incremental tail reads accumulate correctly.
- * Key: session prefix → { inputTokens, outputTokens, ... }
- */
-const telemetryAccumulators = new Map();
-
-/**
- * Tracks which files are currently contributing to each prefix.
- * Key: session prefix → Set of absolute file paths
- * Used to detect file rotation: when files change, accumulator is reset.
- */
-const telemetryPrefixFiles = new Map();
-
-/**
- * Parse a telemetry JSONL filename to extract lane number and role.
- * Pattern: {opId}-{batchId}-{repoId}[-{taskId}][-lane-{N}]-{role}.jsonl
- * Roles: worker, reviewer, merger
- * Returns { laneNumber: number|null, role: string, mergeNumber: number|null } or null if unparseable.
- */
-function parseTelemetryFilename(filename) {
-  // Remove .jsonl extension
-  const base = filename.replace(/\.jsonl$/, "");
-  // Role is always the last segment
-  const lastDash = base.lastIndexOf("-");
-  if (lastDash < 0) return null;
-  const role = base.slice(lastDash + 1);
-  if (role !== "worker" && role !== "reviewer" && role !== "merger") return null;
-
-  // Extract lane number from -lane-{N}- pattern
-  const laneMatch = base.match(/-lane-(\d+)-/);
-  const laneNumber = laneMatch ? parseInt(laneMatch[1], 10) : null;
-
-  // Extract merge number from -merge-{N}- pattern (merge agents)
-  const mergeMatch = base.match(/-merge-(\d+)-/);
-  const mergeNumber = mergeMatch ? parseInt(mergeMatch[1], 10) : null;
-
-  return { laneNumber, role, mergeNumber };
-}
-
-/**
- * Incrementally read new bytes from a JSONL file, parse events, and return them.
- * Handles: file not yet created, empty reads, partial trailing lines, malformed JSON.
- * @param {string} filePath - Absolute path to the JSONL file
- * @returns {object[]} Array of parsed event objects from new data
- */
-function tailJsonlFile(filePath) {
-  // Get or create tail state for this file
-  let tailState = telemetryTailStates.get(filePath);
-  if (!tailState) {
-    tailState = { offset: 0, partial: "" };
-    telemetryTailStates.set(filePath, tailState);
-  }
-
-  // Check file size
-  let fileSize;
-  try {
-    fileSize = fs.statSync(filePath).size;
-  } catch {
-    return []; // File doesn't exist yet
-  }
-
-  // Handle file truncation/recreation (offset beyond current size)
-  if (fileSize < tailState.offset) {
-    tailState.offset = 0;
-    tailState.partial = "";
-    tailState.wasReset = true; // Signal to caller that accumulator should be reset
-  }
-
-  if (fileSize <= tailState.offset) {
-    return []; // No new data
-  }
-
-  // Cap read size per tick to avoid ERR_STRING_TOO_LONG on large files.
-  // If there's more data remaining, the next SSE tick will pick up the rest.
-  const MAX_TAIL_BYTES = 10 * 1024 * 1024; // 10 MB per tick
-
-  // Skip-to-tail on fresh dashboard start with large files.
-  // The partial-line handling below already discards the first partial line.
-  if (tailState.offset === 0 && fileSize > MAX_TAIL_BYTES) {
-    tailState.offset = fileSize - MAX_TAIL_BYTES;
-  }
-
-  // Read new bytes from offset, capped to MAX_TAIL_BYTES
-  const bytesToRead = Math.min(fileSize - tailState.offset, MAX_TAIL_BYTES);
-  const buf = Buffer.alloc(bytesToRead);
-  let fd;
-  try {
-    fd = fs.openSync(filePath, "r");
-  } catch {
-    return []; // File became inaccessible
-  }
-  try {
-    fs.readSync(fd, buf, 0, bytesToRead, tailState.offset);
-  } catch {
-    fs.closeSync(fd);
-    return []; // Read error — try again next tick
-  }
-  fs.closeSync(fd);
-  tailState.offset += bytesToRead;
-
-  // Split into lines, preserving partial trailing line
-  const chunk = tailState.partial + buf.toString("utf-8");
-  const lines = chunk.split("\n");
-  tailState.partial = lines.pop() || "";
-
-  const events = [];
-  for (const line of lines) {
-    const trimmed = line.trim();
-    if (!trimmed) continue;
-    try {
-      const event = JSON.parse(trimmed);
-      if (event && event.type) events.push(event);
-    } catch {
-      // Malformed JSON — skip (concurrent write race, truncated line)
-    }
-  }
-  return events;
-}
-
-/**
- * Load and accumulate telemetry from .pi/telemetry/*.jsonl files.
- * Returns telemetry keyed by session prefix (e.g., "orch-lane-1").
- *
- * Uses batch-state lanes to map lane numbers → session prefixes.
- * For standalone /task mode (no lane number in filename), data is keyed as "standalone".
- *
- * @param {object|null} batchState - The batch state from batch-state.json
- * @returns {object} Map of sessionPrefix → accumulated telemetry
- */
-
-// ── Runtime V2 Data Loaders (TP-107) ─────────────────────────────
-
-/**
- * Load the Runtime V2 process registry for the current batch.
- * Returns null if no registry exists (legacy batch).
- */
-function loadRuntimeRegistry(batchId) {
-  if (!batchId) return null;
-  const registryPath = path.join(REPO_ROOT, ".pi", "runtime", batchId, "registry.json");
-  try {
-    if (!fs.existsSync(registryPath)) return null;
-    return JSON.parse(fs.readFileSync(registryPath, "utf-8"));
-  } catch {
-    return null;
-  }
-}
-
-/**
- * Load Runtime V2 lane snapshots for the current batch.
- * Returns a map of laneNumber → snapshot data.
- */
-function loadRuntimeLaneSnapshots(batchId) {
-  if (!batchId) return {};
-  const lanesDir = path.join(REPO_ROOT, ".pi", "runtime", batchId, "lanes");
-  const snapshots = {};
-  try {
-    if (!fs.existsSync(lanesDir)) return snapshots;
-    const files = fs.readdirSync(lanesDir).filter(f => f.startsWith("lane-") && f.endsWith(".json"));
-    for (const file of files) {
-      try {
-        const data = JSON.parse(fs.readFileSync(path.join(lanesDir, file), "utf-8"));
-        if (data.laneNumber != null) snapshots[data.laneNumber] = data;
-      } catch { continue; }
-    }
-  } catch { /* dir missing */ }
-  return snapshots;
-}
-
-/**
- * Load Runtime V2 merge agent snapshots for the current batch.
- *
- * Reads all `merge-N.json` files from `.pi/runtime/{batchId}/lanes/`.
- * Returns a map of mergeNumber (string) → snapshot data.
- *
- * Follows the same pattern as {@link loadRuntimeLaneSnapshots}.
- *
- * @since TP-164
- */
-function loadRuntimeMergeSnapshots(batchId) {
-  if (!batchId) return {};
-  const lanesDir = path.join(REPO_ROOT, ".pi", "runtime", batchId, "lanes");
-  const snapshots = {};
-  try {
-    if (!fs.existsSync(lanesDir)) return snapshots;
-    const files = fs.readdirSync(lanesDir).filter(f => f.startsWith("merge-") && f.endsWith(".json"));
-    for (const file of files) {
-      try {
-        const data = JSON.parse(fs.readFileSync(path.join(lanesDir, file), "utf-8"));
-        if (data.mergeNumber != null) snapshots[data.mergeNumber] = data;
-      } catch { continue; }
-    }
-  } catch { /* dir missing */ }
-  return snapshots;
-}
-
-/**
- * Load Runtime V2 agent events for a specific agent.
- * Returns the last N events from the agent's events.jsonl.
- */
-function loadRuntimeAgentEvents(batchId, agentId, maxEvents) {
-  if (!batchId || !agentId) return [];
-  maxEvents = maxEvents || 200;
-  const eventsPath = path.join(REPO_ROOT, ".pi", "runtime", batchId, "agents", agentId, "events.jsonl");
-  try {
-    if (!fs.existsSync(eventsPath)) return [];
-    const raw = fs.readFileSync(eventsPath, "utf-8");
-    const lines = raw.split("\n").filter(l => l.trim());
-    const events = [];
-    const start = Math.max(0, lines.length - maxEvents);
-    for (let i = start; i < lines.length; i++) {
-      try { events.push(JSON.parse(lines[i])); } catch { continue; }
-    }
-    return events;
-  } catch {
-    return [];
-  }
-}
-
-/**
- * Load mailbox message activity for the current batch.
- *
- * TP-093 hardening: event-authoritative model.
- * Primary source: .pi/mailbox/{batchId}/events.jsonl (audit event stream).
- * Fallback: directory scans (inbox/ack/outbox/outbox/processed) for
- * compatibility when events.jsonl is absent.
- *
- * Includes:
- * - Consumed replies (outbox/processed/) so they don't disappear after ack
- * - Per-recipient broadcast delivery state from ack markers
- * - Rate-limited events in the timeline
- */
-function loadMailboxData(batchId) {
-  if (!batchId) return { messages: [], agentIds: [], auditEvents: [] };
-  const mbRoot = path.join(REPO_ROOT, ".pi", "mailbox", batchId);
-  if (!fs.existsSync(mbRoot)) return { messages: [], agentIds: [], auditEvents: [] };
-
-  // ── Primary: events.jsonl audit trail ──
-  const auditEvents = loadMailboxAuditEvents(mbRoot);
-
-  // ── Fallback: directory scan ──
-  const messages = [];
-  const agentIds = [];
-
-  try {
-    const dirs = fs.readdirSync(mbRoot, { withFileTypes: true })
-      .filter(d => d.isDirectory())
-      .map(d => d.name);
-
-    for (const agentDir of dirs) {
-      if (agentDir === "_broadcast") continue;
-      agentIds.push(agentDir);
-
-      // Scan inbox (pending), ack (delivered), outbox (active replies), outbox/processed (consumed replies)
-      for (const subdir of ["inbox", "ack", "outbox", "outbox/processed"]) {
-        const dir = path.join(mbRoot, agentDir, subdir);
-        if (!fs.existsSync(dir)) continue;
-        try {
-          const files = fs.readdirSync(dir).filter(f => f.endsWith(".msg.json"));
-          for (const file of files) {
-            try {
-              const msg = JSON.parse(fs.readFileSync(path.join(dir, file), "utf-8"));
-              let status;
-              if (subdir === "inbox") status = "pending";
-              else if (subdir === "ack") status = "delivered";
-              else if (subdir === "outbox") status = "reply";
-              else status = "reply-acked"; // outbox/processed
-              const isBroadcast = msg.to === "_broadcast";
-              messages.push({ ...msg, _status: status, _agentDir: agentDir, _isBroadcast: isBroadcast });
-            } catch { continue; }
-          }
-        } catch { continue; }
-      }
-    }
-
-    // _broadcast: per-recipient delivery state
-    const broadcastInbox = path.join(mbRoot, "_broadcast", "inbox");
-    const broadcastAck = path.join(mbRoot, "_broadcast", "ack");
-    for (const [dir, status] of [[broadcastInbox, "pending"], [broadcastAck, "delivered"]]) {
-      if (!fs.existsSync(dir)) continue;
-      try {
-        const files = fs.readdirSync(dir).filter(f => f.endsWith(".msg.json"));
-        for (const file of files) {
-          try {
-            const msg = JSON.parse(fs.readFileSync(path.join(dir, file), "utf-8"));
-            messages.push({ ...msg, _status: status, _agentDir: "_broadcast", _isBroadcast: true });
-          } catch { continue; }
-        }
-      } catch { continue; }
-    }
-  } catch { /* mailbox dir issues */ }
-
-  messages.sort((a, b) => (a.timestamp || 0) - (b.timestamp || 0));
-  return { messages, agentIds, auditEvents };
-}
-
-/**
- * Load mailbox audit events from events.jsonl.
- * Returns events sorted by timestamp. Includes: message_sent, message_delivered,
- * message_replied, message_escalated, message_rate_limited.
- */
-function loadMailboxAuditEvents(mbRoot) {
-  const eventsPath = path.join(mbRoot, "events.jsonl");
-  if (!fs.existsSync(eventsPath)) return [];
-  try {
-    const raw = fs.readFileSync(eventsPath, "utf-8");
-    const events = [];
-    for (const line of raw.split("\n")) {
-      if (!line.trim()) continue;
-      try { events.push(JSON.parse(line)); } catch { continue; }
-    }
-    return events;
-  } catch {
-    return [];
-  }
-}
-
-// ─────────────────────────────────────────────────────────────────────
-function loadTelemetryData(batchState) {
-  const telemetryDir = path.join(REPO_ROOT, ".pi", "telemetry");
-  const result = {};
-
-  // Build lane number → session prefix mapping from batch state
-  const laneToPrefix = {};
-  if (batchState && batchState.lanes) {
-    for (const lane of batchState.lanes) {
-      const laneSessionId = lane.laneSessionId;
-      if (lane.laneNumber != null && laneSessionId) {
-        laneToPrefix[lane.laneNumber] = laneSessionId;
-      }
-    }
-  }
-
-  // Scan telemetry directory for JSONL files
-  let files;
-  try {
-    files = fs.readdirSync(telemetryDir).filter(f => f.endsWith(".jsonl"));
-  } catch {
-    // .pi/telemetry/ may not exist (pre-RPC sessions) — degrade gracefully
-    return result;
-  }
-
-  // Track which files still exist for tail-state cleanup
-  const currentFiles = new Set();
-  // Track current file→prefix mapping to detect file rotation
-  const currentPrefixFiles = new Map(); // prefix → Set<filePath>
-
-  for (const file of files) {
-    const filePath = path.join(telemetryDir, file);
-    currentFiles.add(filePath);
-
-    // Parse filename to get lane number and role
-    const parsed = parseTelemetryFilename(file);
-    if (!parsed) continue;
-
-    // Determine the key (session prefix)
-    let prefix;
-    if (parsed.role === "merger") {
-      // Merge agent — derive prefix from lane naming so it matches the
-      // session ID used by the client (e.g. "orch-henrylach-merge-1").
-      // Lane sessions: "orch-{opId}-lane-{N}" → merge sessions: "orch-{opId}-merge-{N}".
-      const firstLanePrefix = Object.values(laneToPrefix)[0]; // e.g. "orch-henrylach-lane-1"
-      const opPrefix = firstLanePrefix?.replace(/-lane-\d+$/, ""); // "orch-henrylach"
-      if (parsed.mergeNumber != null && opPrefix) {
-        prefix = `${opPrefix}-merge-${parsed.mergeNumber}`;
-      } else if (parsed.mergeNumber != null) {
-        prefix = `orch-merge-${parsed.mergeNumber}`;
-      } else {
-        prefix = "orch-merge";
-      }
-    } else if (parsed.laneNumber != null && laneToPrefix[parsed.laneNumber]) {
-      prefix = laneToPrefix[parsed.laneNumber];
-    } else if (parsed.laneNumber != null) {
-      // Lane number found but no batch-state mapping — use heuristic
-      prefix = `orch-lane-${parsed.laneNumber}`;
-    } else {
-      // Standalone /task mode
-      prefix = "standalone";
-    }
-
-    // Track file→prefix mapping
-    if (!currentPrefixFiles.has(prefix)) currentPrefixFiles.set(prefix, new Set());
-    currentPrefixFiles.get(prefix).add(filePath);
-
-    // Check if file set for this prefix has changed (file rotation)
-    const prevFiles = telemetryPrefixFiles.get(prefix);
-    const isNewFile = !prevFiles || !prevFiles.has(filePath);
-
-    // Initialize persistent accumulator for this prefix if needed,
-    // or reset if files changed (new file appeared for same prefix)
-    if (!telemetryAccumulators.has(prefix) || (isNewFile && !telemetryTailStates.has(filePath))) {
-      const fresh = {
-        inputTokens: 0, outputTokens: 0, cacheReadTokens: 0,
-        cacheWriteTokens: 0, cost: 0, toolCalls: 0,
-        lastTool: "", currentTool: "", retries: 0, retryActive: false,
-        lastRetryError: "", compactions: 0, latestTotalTokens: 0,
-        contextPct: 0, startedAt: 0,
-      };
-      telemetryAccumulators.set(prefix, fresh);
-      // Also reset tail states for ALL files of this prefix to re-read from beginning
-      if (prevFiles) {
-        for (const pf of prevFiles) {
-          telemetryTailStates.delete(pf);
-        }
-      }
-    }
-
-    const acc = telemetryAccumulators.get(prefix);
-    result[prefix] = acc; // expose the persistent accumulator in the result
-
-    // Tail the file for new events
-    const events = tailJsonlFile(filePath);
-
-    // Check if file was truncated — reset accumulator
-    const ts = telemetryTailStates.get(filePath);
-    if (ts && ts.wasReset) {
-      acc.inputTokens = 0; acc.outputTokens = 0; acc.cacheReadTokens = 0;
-      acc.cacheWriteTokens = 0; acc.cost = 0; acc.toolCalls = 0;
-      acc.lastTool = ""; acc.currentTool = ""; acc.retries = 0; acc.retryActive = false;
-      acc.lastRetryError = ""; acc.compactions = 0; acc.latestTotalTokens = 0;
-      acc.contextPct = 0; acc.startedAt = 0;
-      ts.wasReset = false;
-    }
-    for (const event of events) {
-      switch (event.type) {
-        case "message_end": {
-          // A successful message_end means any prior retry resolved.
-          // Clear retryActive to prevent stale retry badges from persisting
-          // across batches or after transient API errors recover.
-          acc.retryActive = false;
-          const usage = event.message?.usage;
-          if (usage) {
-            acc.inputTokens += usage.input || 0;
-            acc.outputTokens += usage.output || 0;
-            acc.cacheReadTokens += usage.cacheRead || 0;
-            acc.cacheWriteTokens += usage.cacheWrite || 0;
-            if (usage.cost) {
-              acc.cost += typeof usage.cost === "object"
-                ? (usage.cost.total || 0)
-                : (typeof usage.cost === "number" ? usage.cost : 0);
-            }
-            // Include cacheRead: totalTokens from pi excludes cache reads,
-            // but cached tokens still consume context window capacity.
-            const rawTotal = usage.totalTokens
-              || ((usage.input || 0) + (usage.output || 0));
-            const totalTokens = rawTotal + (usage.cacheRead || 0);
-            if (totalTokens > acc.latestTotalTokens) {
-              acc.latestTotalTokens = totalTokens;
-            }
-          }
-          break;
-        }
-        case "tool_execution_start": {
-          acc.toolCalls++;
-          const toolDesc = event.toolName || "unknown";
-          let argPreview = "";
-          if (event.args) {
-            if (typeof event.args === "string") {
-              argPreview = event.args.slice(0, 80);
-            } else if (typeof event.args === "object") {
-              const firstVal = Object.values(event.args)[0];
-              if (typeof firstVal === "string") {
-                argPreview = firstVal.slice(0, 80);
-              }
-            }
-          }
-          const toolLabel = argPreview ? `${toolDesc} ${argPreview}` : toolDesc;
-          acc.lastTool = toolLabel;
-          acc.currentTool = toolLabel;
-          break;
-        }
-        case "tool_execution_end": {
-          acc.currentTool = "";
-          break;
-        }
-        case "agent_start": {
-          if (event.ts && !acc.startedAt) {
-            acc.startedAt = event.ts;
-          }
-          break;
-        }
-        case "response": {
-          // Extract context usage from get_session_stats responses
-          const ctxUsage = event.data?.contextUsage;
-          if (ctxUsage) {
-            // Support both percent (current) and percentUsed (legacy pi versions)
-            const pct = typeof ctxUsage.percent === "number" ? ctxUsage.percent
-              : typeof ctxUsage.percentUsed === "number" ? ctxUsage.percentUsed
-              : null;
-            if (pct !== null) acc.contextPct = pct;
-          }
-          break;
-        }
-        case "auto_retry_start": {
-          acc.retries++;
-          acc.retryActive = true;
-          acc.lastRetryError = event.errorMessage || event.error || "unknown";
-          break;
-        }
-        case "auto_retry_end": {
-          acc.retryActive = false;
-          break;
-        }
-        case "auto_compaction_start": {
-          acc.compactions++;
-          break;
-        }
-      }
-    }
-  }
-
-  // Clean up tail states for files that no longer exist
-  for (const [filePath] of telemetryTailStates) {
-    if (filePath.startsWith(telemetryDir) && !currentFiles.has(filePath)) {
-      telemetryTailStates.delete(filePath);
-    }
-  }
-
-  // Update prefix→files tracking for next call
-  // Clean up accumulators and tracking for prefixes that have no remaining files
-  const activePrefixes = new Set(Object.keys(result));
-  for (const [prefix] of telemetryAccumulators) {
-    if (!activePrefixes.has(prefix)) {
-      telemetryAccumulators.delete(prefix);
-      telemetryPrefixFiles.delete(prefix);
-    }
-  }
-  // Store current file mappings for next call's rotation detection
-  for (const [prefix, fileSet] of currentPrefixFiles) {
-    telemetryPrefixFiles.set(prefix, fileSet);
-  }
-
-  return result;
-}
-
-// ─── Supervisor Data Loading ────────────────────────────────────────────────
-
-/**
- * Module-level tail state for supervisor JSONL files (actions.jsonl, events.jsonl).
- * Reuses the same incremental tailing pattern as telemetry.
- * Key: absolute file path → { offset, partial, entries }
- */
-const supervisorTailStates = {
-  actions: { offset: 0, partial: "", entries: [] },
-  events: { offset: 0, partial: "", entries: [] },
-  conversation: { offset: 0, partial: "", entries: [] },
-};
-
-/**
- * The last known batchId — used to detect batch changes and reset accumulators.
- */
-let supervisorLastBatchId = "";
-
-/**
- * Incrementally tail a JSONL file, accumulating parsed entries.
- * Filters entries by batchId when provided.
- *
- * @param {string} filePath - Absolute path to the JSONL file
- * @param {object} tailState - Mutable tail state { offset, partial, entries }
- * @param {string} batchId - Batch ID to filter by (empty = no filter)
- * @returns {object[]} The accumulated entries array (same reference as tailState.entries)
- */
-function tailSupervisorJsonl(filePath, tailState, batchId) {
-  // Check file size
-  let fileSize;
-  try {
-    fileSize = fs.statSync(filePath).size;
-  } catch {
-    return tailState.entries; // File doesn't exist yet — return accumulated
-  }
-
-  // Handle file truncation/recreation
-  if (fileSize < tailState.offset) {
-    tailState.offset = 0;
-    tailState.partial = "";
-    tailState.entries = [];
-  }
-
-  if (fileSize <= tailState.offset) {
-    return tailState.entries; // No new data
-  }
-
-  // Read new bytes from offset
-  const bytesToRead = fileSize - tailState.offset;
-  const buf = Buffer.alloc(bytesToRead);
-  let fd;
-  try {
-    fd = fs.openSync(filePath, "r");
-  } catch {
-    return tailState.entries;
-  }
-  try {
-    fs.readSync(fd, buf, 0, bytesToRead, tailState.offset);
-  } catch {
-    fs.closeSync(fd);
-    return tailState.entries;
-  }
-  fs.closeSync(fd);
-  tailState.offset = fileSize;
-
-  // Split into lines, preserving partial trailing line
-  const chunk = tailState.partial + buf.toString("utf-8");
-  const lines = chunk.split("\n");
-  tailState.partial = lines.pop() || "";
-
-  for (const line of lines) {
-    const trimmed = line.trim();
-    if (!trimmed) continue;
-    try {
-      const entry = JSON.parse(trimmed);
-      // Filter by batchId if provided
-      if (batchId && entry.batchId && entry.batchId !== batchId) continue;
-      tailState.entries.push(entry);
-    } catch {
-      // Malformed JSON — skip
-    }
-  }
-
-  // Cap accumulated entries to prevent unbounded growth (keep last 500)
-  if (tailState.entries.length > 500) {
-    tailState.entries = tailState.entries.slice(-500);
-  }
-
-  return tailState.entries;
-}
-
-/**
- * Read supervisor autonomy level from project config.
- *
- * Checks `.pi/taskplane-config.json` for `orchestrator.supervisor.autonomy`.
- * Falls back to "supervised" (the default) if config is missing or malformed.
- * This is needed because the lockfile does not contain the autonomy level.
- *
- * @returns {string} Autonomy level: "interactive" | "supervised" | "autonomous"
- */
-function loadSupervisorAutonomy() {
-  try {
-    const configPath = path.join(REPO_ROOT, ".pi", "taskplane-config.json");
-    const raw = fs.readFileSync(configPath, "utf-8");
-    const config = JSON.parse(raw);
-    const autonomy = config?.orchestrator?.supervisor?.autonomy;
-    if (autonomy === "interactive" || autonomy === "supervised" || autonomy === "autonomous") {
-      return autonomy;
-    }
-  } catch {
-    // Config missing or malformed — use default
-  }
-  return "supervised"; // Default per DEFAULT_SUPERVISOR_CONFIG
-}
-
-/**
- * Load supervisor data for the dashboard.
- *
- * Reads (all from .pi/supervisor/):
- * - lock.json: supervisor active/stale status, heartbeat, autonomy (from config)
- * - actions.jsonl: recovery action audit trail (batch-scoped, incremental)
- * - events.jsonl: engine + tier 0 events (batch-scoped, incremental)
- * - conversation.jsonl: operator ↔ supervisor interaction log (spec §9.1)
- * - summary.md: human-readable batch summary (generated on completion)
- *
- * Returns null when no supervisor files exist (pre-supervisor batches).
- *
- * @param {object|null} batchState - The batch state from batch-state.json
- * @returns {object|null} Supervisor data object or null
- */
-function loadSupervisorData(batchState) {
-  const supervisorDir = path.join(REPO_ROOT, ".pi", "supervisor");
-  const batchId = batchState ? (batchState.batchId || "") : "";
-
-  // Detect batch change — reset tail state accumulators
-  if (batchId && batchId !== supervisorLastBatchId) {
-    supervisorLastBatchId = batchId;
-    supervisorTailStates.actions = { offset: 0, partial: "", entries: [] };
-    supervisorTailStates.events = { offset: 0, partial: "", entries: [] };
-    supervisorTailStates.conversation = { offset: 0, partial: "", entries: [] };
-  }
-
-  // ── Lockfile: supervisor status ──
-  // The lockfile contains pid, sessionId, batchId, startedAt, heartbeat.
-  // It does NOT contain autonomy — that comes from project config.
-  let lock = null;
-  try {
-    const lockPath = path.join(supervisorDir, "lock.json");
-    const raw = fs.readFileSync(lockPath, "utf-8");
-    const parsed = JSON.parse(raw);
-    if (parsed && parsed.pid && parsed.sessionId) {
-      // Determine if lock is stale (heartbeat older than 90s)
-      const heartbeatAge = parsed.heartbeat
-        ? Date.now() - new Date(parsed.heartbeat).getTime()
-        : Infinity;
-      const isStale = heartbeatAge > 90_000;
-
-      lock = {
-        active: !isStale,
-        pid: parsed.pid,
-        sessionId: parsed.sessionId,
-        batchId: parsed.batchId || "",
-        startedAt: parsed.startedAt || "",
-        heartbeat: parsed.heartbeat || "",
-        // Autonomy is NOT in the lockfile — derive from project config
-        autonomy: loadSupervisorAutonomy(),
-      };
-    }
-  } catch {
-    // No lockfile or malformed — supervisor is inactive
-  }
-
-  // ── Actions JSONL: recovery audit trail (batch-scoped, incremental) ──
-  const actionsPath = path.join(supervisorDir, "actions.jsonl");
-  const actions = tailSupervisorJsonl(actionsPath, supervisorTailStates.actions, batchId);
-
-  // ── Events JSONL: engine events (batch-scoped, incremental) ──
-  const eventsPath = path.join(supervisorDir, "events.jsonl");
-  const events = tailSupervisorJsonl(eventsPath, supervisorTailStates.events, batchId);
-
-  // ── Conversation JSONL: operator interaction log (spec §9.1) ──
-  // The supervisor writes operator↔supervisor messages to conversation.jsonl.
-  // Not yet implemented in all supervisor versions — degrade gracefully.
-  const conversationPath = path.join(supervisorDir, "conversation.jsonl");
-  const conversation = tailSupervisorJsonl(conversationPath, supervisorTailStates.conversation, batchId);
-
-  // ── Summary: human-readable batch summary (generated on completion) ──
-  // Per spec §9.1, the supervisor writes .pi/supervisor/summary.md when the
-  // batch completes or is abandoned. Read the file if it exists.
-  let summary = null;
-  try {
-    const summaryPath = path.join(supervisorDir, "summary.md");
-    summary = fs.readFileSync(summaryPath, "utf-8");
-  } catch {
-    // No summary yet — batch may still be running, or pre-supervisor batch
-  }
-
-  // If nothing exists at all, return null (pre-supervisor batch)
-  if (!lock && actions.length === 0 && events.length === 0 && conversation.length === 0 && !summary) {
-    // Check if the supervisor directory even exists
-    try {
-      fs.statSync(supervisorDir);
-    } catch {
-      return null; // No supervisor dir → pre-supervisor batch
-    }
-  }
-
-  return {
-    lock,
-    actions,
-    events,
-    conversation,
-    summary,
-  };
-}
-
-/**
- * Compute batch total cost from lane states (primary) and telemetry (supplementary).
- * Lane states are authoritative — telemetry provides additional data only for lanes
- * that have no lane-state entry (e.g., very early in session startup).
- */
-function computeBatchTotalCost(laneStates, telemetry) {
-  let totalCost = 0;
-  const coveredPrefixes = new Set();
-
-  // Primary: sum cost from lane states (worker + reviewer)
-  for (const [prefix, ls] of Object.entries(laneStates)) {
-    if (ls.workerCostUsd) {
-      totalCost += ls.workerCostUsd;
-      coveredPrefixes.add(prefix);
-    }
-    if (ls.reviewerCostUsd) {
-      totalCost += ls.reviewerCostUsd;
-    }
-  }
-
-  // Supplementary: add cost from telemetry for uncovered lanes only
-  for (const [prefix, tel] of Object.entries(telemetry)) {
-    if (!coveredPrefixes.has(prefix) && tel.cost > 0) {
-      totalCost += tel.cost;
-    }
-  }
-
-  return totalCost;
-}
-
-function synthesizeLaneStateFromSnapshot(key, snap, fallbackBatchId) {
-  const w = snap.worker || {};
-  const r = snap.reviewer || null;
-  const statusMap = { running: "running", spawning: "running", exited: "done", crashed: "error", killed: "error", timed_out: "error", wrapping_up: "running" };
-  const reviewerStatusMap = { running: "running", spawning: "running", wrapping_up: "running", exited: "done", crashed: "done", killed: "done", timed_out: "done" };
-
-  return {
-    prefix: key,
-    taskId: snap.taskId || null,
-    phase: snap.status === "running" ? "worker-active" : snap.status === "complete" ? "complete" : "idle",
-    workerStatus: statusMap[w.status] || w.status || "idle",
-    workerElapsed: w.elapsedMs || 0,
-    workerContextPct: w.contextPct || 0,
-    workerLastTool: w.lastTool || "",
-    workerToolCount: w.toolCalls || 0,
-    workerInputTokens: w.inputTokens || 0,
-    workerOutputTokens: w.outputTokens || 0,
-    workerCacheReadTokens: w.cacheReadTokens || 0,
-    workerCacheWriteTokens: w.cacheWriteTokens || 0,
-    workerCostUsd: w.costUsd || 0,
-    reviewerStatus: r ? (reviewerStatusMap[r.status] || r.status || "running") : "idle",
-    reviewerElapsed: r?.elapsedMs || 0,
-    reviewerContextPct: r?.contextPct || 0,
-    reviewerLastTool: r?.lastTool || "",
-    reviewerToolCount: r?.toolCalls || 0,
-    reviewerCostUsd: r?.costUsd || 0,
-    reviewerInputTokens: r?.inputTokens || 0,
-    reviewerOutputTokens: r?.outputTokens || 0,
-    reviewerCacheReadTokens: r?.cacheReadTokens || 0,
-    reviewerCacheWriteTokens: r?.cacheWriteTokens || 0,
-    reviewerType: r?.reviewType || "",
-    reviewerStep: r?.reviewStep || 0,
-    batchId: snap.batchId || fallbackBatchId,
-    timestamp: snap.updatedAt || Date.now(),
-  };
-}
-
-/** Build full dashboard state object for the frontend. */
-function buildDashboardState() {
-  const state = loadBatchState();
-  const sessions = getActiveSessions();
-  const rawLaneStates = loadLaneStates();
-  // Filter stale lane states from previous batches.
-  // Lane state files persist across batches (same filename), so without
-  // filtering the dashboard shows telemetry from old runs.
-  const currentBatchId = state?.batchId || null;
-  const laneStates = {};
-  for (const [key, ls] of Object.entries(rawLaneStates)) {
-    if (!currentBatchId || !ls.batchId || ls.batchId === currentBatchId) {
-      laneStates[key] = ls;
-    }
-  }
-  const telemetry = loadTelemetryData(state);
-  const batchTotalCost = computeBatchTotalCost(laneStates, telemetry);
-  const supervisor = loadSupervisorData(state);
-
-  if (!state) {
-    return {
-      batch: null,
-      sessions,
-      tmuxSessions: sessions, // Legacy compatibility field for older dashboard clients
-      laneStates: {},
-      telemetry: {},
-      batchTotalCost: 0,
-      supervisor: null,
-      timestamp: Date.now(),
-    };
-  }
-
-  const tasks = (state.tasks || []).map((task) => {
-    const effectiveFolder = resolveTaskFolder(task, state);
-    let statusData = null;
-    if (effectiveFolder) {
-      statusData = parseStatusMd(effectiveFolder);
-    }
-    if (!task.doneFileFound && effectiveFolder) {
-      task.doneFileFound = checkDoneFile(effectiveFolder);
-    }
-    return { ...task, statusData };
-  });
-
-  // TP-107: Load Runtime V2 data when available
-  const runtimeRegistry = loadRuntimeRegistry(state.batchId);
-  const runtimeLaneSnapshots = loadRuntimeLaneSnapshots(state.batchId);
-  const mailboxData = loadMailboxData(state.batchId);
-
-  // TP-164: Load merge agent snapshots for live dashboard telemetry.
-  const runtimeMergeSnapshots = loadRuntimeMergeSnapshots(state.batchId);
-
-  // TP-164: Inject merge snapshot telemetry into the telemetry map so
-  // `telemetry[sessionName]` resolves for the merge pane.
-  // Snapshots are the PRIMARY source for merge telemetry — merge agents don't
-  // write to .pi/telemetry/*.jsonl (the JSONL path is for lane workers only).
-  // Inject snapshot data when no JSONL-backed entry exists for this session.
-  for (const snap of Object.values(runtimeMergeSnapshots)) {
-    const key = snap.sessionName;
-    if (!key) continue;
-    const agent = snap.agent;
-    if (!agent) continue;
-    // Only inject if there is no existing JSONL-backed telemetry entry.
-    // Merge agents don't emit to .pi/telemetry/, so existing will always be
-    // absent unless something else wrote to it — in that case defer to it.
-    const existing = telemetry[key];
-    if (!existing) {
-      telemetry[key] = {
-        inputTokens: agent.inputTokens || 0,
-        outputTokens: agent.outputTokens || 0,
-        cacheReadTokens: agent.cacheReadTokens || 0,
-        cacheWriteTokens: agent.cacheWriteTokens || 0,
-        cost: agent.costUsd || 0,
-        toolCalls: agent.toolCalls || 0,
-        lastTool: agent.lastTool || "",
-        currentTool: snap.status === "running" ? (agent.lastTool || "") : "",
-        contextPct: agent.contextPct || 0,
-        // startedAt is not in the snapshot; compute from elapsed if possible.
-        startedAt: agent.elapsedMs > 0 ? snap.updatedAt - agent.elapsedMs : snap.updatedAt,
-        retries: 0,
-        retryActive: false,
-        lastRetryError: "",
-        compactions: 0,
-        latestTotalTokens: (agent.inputTokens || 0) + (agent.outputTokens || 0),
-        _updatedAt: snap.updatedAt,
-        _source: "merge-snapshot",
-        // TP-178: Include waveIndex for precise wave-telemetry association (#498)
-        waveIndex: snap.waveIndex != null ? snap.waveIndex : undefined,
-      };
-    }
-  }
-
-  // TP-115: Synthesize laneStates from V2 snapshots so the dashboard
-  // pipeline works without legacy lane-state-*.json sidecar files.
-  // V2 snapshots are authoritative when present.
-  if (Object.keys(runtimeLaneSnapshots).length > 0) {
-    for (const [laneNum, snap] of Object.entries(runtimeLaneSnapshots)) {
-      // Find the matching lane record to get the session name key
-      const laneRec = (state.lanes || []).find(l => l.laneNumber === Number(laneNum));
-      const key = laneRec ? (laneRec.laneSessionId) : `lane-${laneNum}`;
-      if (!laneStates[key] || (snap.updatedAt && snap.updatedAt > (laneStates[key].timestamp || 0))) {
-        laneStates[key] = synthesizeLaneStateFromSnapshot(key, snap, state.batchId);
-      }
-    }
-  }
-
-  return {
-    laneStates,
-    telemetry,
-    batchTotalCost,
-    supervisor,
-    // Runtime V2 data (null/empty for legacy batches)
-    runtimeRegistry,
-    runtimeLaneSnapshots,
-    // TP-164: Merge agent snapshots for live dashboard telemetry.
-    runtimeMergeSnapshots,
-    mailbox: mailboxData,
-    batch: {
-      batchId: state.batchId,
-      phase: state.phase,
-      startedAt: state.startedAt,
-      updatedAt: state.updatedAt,
-      currentWaveIndex: state.currentWaveIndex || 0,
-      totalWaves: state.totalWaves || (state.wavePlan ? state.wavePlan.length : 0),
-      wavePlan: state.wavePlan || [],
-      // Lanes already include repoId (string|undefined) from PersistedLaneRecord (v2).
-      lanes: state.lanes || [],
-      // Tasks already include repoId, resolvedRepoId (string|undefined) from PersistedTaskRecord (v2).
-      tasks,
-      mergeResults: state.mergeResults || [],
-      errors: state.errors || [],
-      lastError: state.lastError || null,
-      // Workspace mode: "repo" (default/v1) or "workspace" (v2 multi-repo).
-      // Additive field — absent in v1 state files, frontend must default to "repo".
-      mode: state.mode || "repo",
-      // TP-148: Segment records for wave display context (v4+).
-      // Each record has taskId, segmentId, repoId, status.
-      segments: state.segments || [],
-    },
-    sessions,
-    tmuxSessions: sessions, // Legacy compatibility field for older dashboard clients
-    timestamp: Date.now(),
-  };
-}
-
-// ─── Static File Serving ────────────────────────────────────────────────────
-
-const MIME_TYPES = {
-  ".html": "text/html; charset=utf-8",
-  ".css": "text/css; charset=utf-8",
-  ".js": "application/javascript; charset=utf-8",
-  ".json": "application/json; charset=utf-8",
-  ".svg": "image/svg+xml",
-  ".png": "image/png",
-  ".ico": "image/x-icon",
-};
-
-function serveStatic(req, res) {
-  let filePath = new URL(req.url, "http://localhost").pathname;
-  if (filePath === "/") filePath = "/index.html";
-
-  const fullPath = path.join(PUBLIC_DIR, filePath);
-  // Prevent directory traversal
-  if (!fullPath.startsWith(PUBLIC_DIR)) {
-    res.writeHead(403);
-    res.end("Forbidden");
-    return;
-  }
-
-  const ext = path.extname(fullPath);
-  const contentType = MIME_TYPES[ext] || "application/octet-stream";
-
-  try {
-    const content = fs.readFileSync(fullPath);
-    res.writeHead(200, {
-      "Content-Type": contentType,
-      "Cache-Control": "no-cache, no-store, must-revalidate",
-    });
-    res.end(content);
-  } catch {
-    res.writeHead(404, { "Content-Type": "text/plain" });
-    res.end("Not Found");
-  }
-}
-
-// ─── SSE Stream ─────────────────────────────────────────────────────────────
-
-const sseClients = new Set();
-
-// ─── Conversation JSONL ─────────────────────────────────────────────────
-
-function serveConversation(req, res, prefix) {
-  if (!/^[\w-]+$/.test(prefix)) {
-    res.writeHead(400, { "Content-Type": "text/plain" });
-    res.end("Invalid prefix");
-    return;
-  }
-
-  const filePath = path.join(REPO_ROOT, ".pi", `worker-conversation-${prefix}.jsonl`);
-  try {
-    const content = fs.readFileSync(filePath, "utf-8");
-    res.writeHead(200, {
-      "Content-Type": "application/x-ndjson",
-      "Access-Control-Allow-Origin": "*",
-    });
-    res.end(content);
-  } catch {
-    res.writeHead(200, {
-      "Content-Type": "application/x-ndjson",
-      "Access-Control-Allow-Origin": "*",
-    });
-    res.end(""); // empty — no conversation yet
-  }
-}
-
-// ─── Dashboard SSE ──────────────────────────────────────────────────────
-
-function handleSSE(req, res) {
-  res.writeHead(200, {
-    "Content-Type": "text/event-stream",
-    "Cache-Control": "no-cache",
-    Connection: "keep-alive",
-    "Access-Control-Allow-Origin": "*",
-  });
-
-  // Send initial state immediately
-  const state = buildDashboardState();
-  res.write(`data: ${JSON.stringify(state)}\n\n`);
-
-  sseClients.add(res);
-  req.on("close", () => sseClients.delete(res));
-}
-
-function broadcastState() {
-  if (sseClients.size === 0) return;
-  const state = buildDashboardState();
-  const payload = `data: ${JSON.stringify(state)}\n\n`;
-  for (const client of sseClients) {
-    try {
-      client.write(payload);
-    } catch {
-      sseClients.delete(client);
-    }
-  }
-}
-
-// ─── Batch History API ──────────────────────────────────────────────────────
-
-// BATCH_HISTORY_PATH is initialized in main() alongside REPO_ROOT.
-
-function loadHistory() {
-  try {
-    if (!fs.existsSync(BATCH_HISTORY_PATH)) return [];
-    const raw = fs.readFileSync(BATCH_HISTORY_PATH, "utf-8");
-    return JSON.parse(raw);
-  } catch {
-    return [];
-  }
-}
-
-/** GET /api/history — return list of batch summaries (compact: no per-task detail). */
-function serveHistory(req, res) {
-  const history = loadHistory();
-  // Return compact list for the dropdown (no per-task details)
-  const compact = history.map(h => ({
-    batchId: h.batchId,
-    status: h.status,
-    startedAt: h.startedAt,
-    endedAt: h.endedAt,
-    durationMs: h.durationMs,
-    totalWaves: h.totalWaves,
-    totalTasks: h.totalTasks,
-    succeededTasks: h.succeededTasks,
-    failedTasks: h.failedTasks,
-    tokens: h.tokens,
-  }));
-  res.writeHead(200, {
-    "Content-Type": "application/json",
-    "Access-Control-Allow-Origin": "*",
-  });
-  res.end(JSON.stringify(compact));
-}
-
-/** GET /api/history/:batchId — return full detail for one batch. */
-function serveHistoryEntry(req, res, batchId) {
-  const history = loadHistory();
-  const entry = history.find(h => h.batchId === batchId);
-  if (!entry) {
-    res.writeHead(404, { "Content-Type": "application/json" });
-    res.end(JSON.stringify({ error: "Batch not found" }));
-    return;
-  }
-  res.writeHead(200, {
-    "Content-Type": "application/json",
-    "Access-Control-Allow-Origin": "*",
-  });
-  res.end(JSON.stringify(entry));
-}
-
-/** GET /api/status-md/:taskId — return raw STATUS.md content for a task. */
-function serveStatusMd(req, res, taskId) {
-  if (!/^[\w-]+$/.test(taskId)) {
-    res.writeHead(400, { "Content-Type": "text/plain" });
-    res.end("Invalid task ID");
-    return;
-  }
-
-  const state = loadBatchState();
-  if (!state) {
-    res.writeHead(404, { "Content-Type": "application/json" });
-    res.end(JSON.stringify({ error: "No batch state" }));
-    return;
-  }
-
-  const task = (state.tasks || []).find(t => t.taskId === taskId);
-  if (!task) {
-    res.writeHead(404, { "Content-Type": "application/json" });
-    res.end(JSON.stringify({ error: "Task not found" }));
-    return;
-  }
-
-  const effectiveFolder = resolveTaskFolder(task, state);
-  if (!effectiveFolder) {
-    res.writeHead(404, { "Content-Type": "application/json" });
-    res.end(JSON.stringify({ error: "Cannot resolve task folder" }));
-    return;
-  }
-
-  // Try effective folder, then archive
-  const candidates = [effectiveFolder];
-  const archiveBase = effectiveFolder.replace(/[/\\]tasks[/\\][^/\\]+$/, "/tasks/archive/" + taskId);
-  if (archiveBase !== effectiveFolder) candidates.push(archiveBase);
-
-  for (const folder of candidates) {
-    const statusPath = path.join(folder, "STATUS.md");
-    try {
-      const content = fs.readFileSync(statusPath, "utf-8");
-      res.writeHead(200, {
-        "Content-Type": "text/plain; charset=utf-8",
-        "Access-Control-Allow-Origin": "*",
-      });
-      res.end(content);
-      return;
-    } catch { continue; }
-  }
-
-  res.writeHead(404, { "Content-Type": "application/json" });
-  res.end(JSON.stringify({ error: "STATUS.md not found" }));
-}
-
-// ─── Dashboard Preferences ──────────────────────────────────────────────────
-
-function getPreferencesPath() {
-  return path.join(REPO_ROOT, ".pi", "dashboard-preferences.json");
-}
-
-function handleGetPreferences(req, res) {
-  const prefsPath = getPreferencesPath();
-  let prefs = { theme: "dark" };
-  try {
-    if (fs.existsSync(prefsPath)) {
-      prefs = JSON.parse(fs.readFileSync(prefsPath, "utf8"));
-    }
-  } catch { /* use defaults */ }
-  res.writeHead(200, {
-    "Content-Type": "application/json",
-    "Access-Control-Allow-Origin": "*",
-  });
-  res.end(JSON.stringify(prefs));
-}
-
-function handlePostPreferences(req, res) {
-  let body = "";
-  req.on("data", (chunk) => { body += chunk; });
-  req.on("end", () => {
-    try {
-      const incoming = JSON.parse(body);
-      const prefsPath = getPreferencesPath();
-      let existing = {};
-      try {
-        if (fs.existsSync(prefsPath)) {
-          existing = JSON.parse(fs.readFileSync(prefsPath, "utf8"));
-        }
-      } catch { /* start fresh */ }
-      const merged = { ...existing, ...incoming };
-      const dir = path.dirname(prefsPath);
-      if (!fs.existsSync(dir)) fs.mkdirSync(dir, { recursive: true });
-      fs.writeFileSync(prefsPath, JSON.stringify(merged, null, 2) + "\n");
-      res.writeHead(200, {
-        "Content-Type": "application/json",
-        "Access-Control-Allow-Origin": "*",
-      });
-      res.end(JSON.stringify(merged));
-    } catch (err) {
-      res.writeHead(400, {
-        "Content-Type": "application/json",
-        "Access-Control-Allow-Origin": "*",
-      });
-      res.end(JSON.stringify({ error: "Invalid JSON" }));
-    }
-  });
-}
-
-// ─── HTTP Server ────────────────────────────────────────────────────────────
-
-function createServer() {
-  const server = http.createServer((req, res) => {
-    const pathname = new URL(req.url, "http://localhost").pathname;
-
-    if (pathname === "/api/stream" && req.method === "GET") {
-      handleSSE(req, res);
-    } else if (pathname.startsWith("/api/conversation/") && req.method === "GET") {
-      const prefix = pathname.slice("/api/conversation/".length);
-      serveConversation(req, res, prefix);
-    } else if (pathname.startsWith("/api/agent-events/") && req.method === "GET") {
-      // TP-107: Serve Runtime V2 agent events (hardened)
-      const agentId = decodeURIComponent(pathname.slice("/api/agent-events/".length));
-      // Strict validation: same pattern as /api/conversation/:prefix
-      if (!/^[\w-]+$/.test(agentId)) {
-        res.writeHead(400, { "Content-Type": "text/plain" });
-        res.end("Invalid agent ID");
-        return;
-      }
-      const batchState = loadBatchState();
-      // Path containment: verify resolved path stays inside runtime dir
-      if (batchState?.batchId) {
-        const runtimeBase = path.join(REPO_ROOT, ".pi", "runtime", batchState.batchId, "agents");
-        const resolvedAgent = path.resolve(runtimeBase, agentId);
-        if (!resolvedAgent.startsWith(path.resolve(runtimeBase))) {
-          res.writeHead(403, { "Content-Type": "text/plain" });
-          res.end("Forbidden");
-          return;
-        }
-      }
-      // Optional: ?sinceTs= to return only events after a timestamp
-      const reqUrl = new URL(req.url, "http://localhost");
-      const sinceTs = parseInt(reqUrl.searchParams.get("sinceTs") || "0", 10);
-      let events = loadRuntimeAgentEvents(batchState?.batchId, agentId, 300);
-      if (sinceTs > 0) {
-        events = events.filter(e => (e.ts || 0) > sinceTs);
-      }
-      res.writeHead(200, { "Content-Type": "application/json", "Access-Control-Allow-Origin": "*" });
-      res.end(JSON.stringify(events));
-    } else if (pathname === "/api/state" && req.method === "GET") {
-      const state = buildDashboardState();
-      res.writeHead(200, {
-        "Content-Type": "application/json",
-        "Access-Control-Allow-Origin": "*",
-      });
-      res.end(JSON.stringify(state));
-    } else if (pathname === "/api/history" && req.method === "GET") {
-      serveHistory(req, res);
-    } else if (pathname.startsWith("/api/history/") && req.method === "GET") {
-      const batchId = decodeURIComponent(pathname.slice("/api/history/".length));
-      serveHistoryEntry(req, res, batchId);
-    } else if (pathname.startsWith("/api/status-md/") && req.method === "GET") {
-      const taskId = decodeURIComponent(pathname.slice("/api/status-md/".length));
-      serveStatusMd(req, res, taskId);
-    } else if (pathname === "/api/preferences" && req.method === "GET") {
-      handleGetPreferences(req, res);
-    } else if (pathname === "/api/preferences" && req.method === "POST") {
-      handlePostPreferences(req, res);
-    } else if (req.method === "OPTIONS") {
-      // CORS preflight for POST
-      res.writeHead(204, {
-        "Access-Control-Allow-Origin": "*",
-        "Access-Control-Allow-Methods": "GET, POST, OPTIONS",
-        "Access-Control-Allow-Headers": "Content-Type",
-      });
-      res.end();
-    } else {
-      serveStatic(req, res);
-    }
-  });
-
-  return server;
-}
-
-// ─── Browser Auto-Open ─────────────────────────────────────────────────────
-
-function openBrowser(url) {
-  const cmd = process.platform === "win32" ? "start"
-    : process.platform === "darwin" ? "open" : "xdg-open";
-  exec(`${cmd} ${url}`, () => {}); // fire-and-forget
-}
-
-// ─── Main ───────────────────────────────────────────────────────────────────
-
-/** Try to listen on a port. Resolves with the port on success, rejects on EADDRINUSE. */
-function tryListen(server, port) {
-  return new Promise((resolve, reject) => {
-    const onError = (err) => {
-      server.removeListener("listening", onListening);
-      reject(err);
-    };
-    const onListening = () => {
-      server.removeListener("error", onError);
-      resolve(port);
-    };
-    server.once("error", onError);
-    server.once("listening", onListening);
-    server.listen(port);
-  });
-}
-
-/** Find an available port starting from `start`, trying up to MAX_PORT_ATTEMPTS. */
-async function findPort(server, start, explicit) {
-  // If the user explicitly passed --port, only try that one
-  if (explicit) {
-    try {
-      return await tryListen(server, start);
-    } catch (err) {
-      if (err.code === "EADDRINUSE") {
-        console.error(`\n  Port ${start} is already in use.`);
-        console.error(`  Try: taskplane dashboard --port ${start + 1}\n`);
-        process.exit(1);
-      }
-      throw err;
-    }
-  }
-  // Auto-scan for an available port
-  for (let port = start; port < start + MAX_PORT_ATTEMPTS; port++) {
-    try {
-      return await tryListen(server, port);
-    } catch (err) {
-      if (err.code === "EADDRINUSE") {
-        // Close the server so we can retry on the next port
-        server.close();
-        server = createServer();
-        continue;
-      }
-      throw err;
-    }
-  }
-  console.error(`\n  No available port found in range ${start}-${start + MAX_PORT_ATTEMPTS - 1}.\n`);
-  process.exit(1);
-}
-
-async function main() {
-  const opts = parseArgs();
-
-  // Resolve project root: --root flag > cwd.
-  // In workspace mode this is the workspace root. All state/sidecar files
-  // (batch-state, lane-state, conversation logs, batch-history) live at
-  // <REPO_ROOT>/.pi/ and are NOT affected by taskplane-pointer.json.
-  // The pointer only redirects config/agent resolution in task-runner and
-  // orchestrator — the dashboard reads only runtime state, so no pointer
-  // resolution is performed here.
-  REPO_ROOT = path.resolve(opts.root || process.cwd());
-  BATCH_STATE_PATH = path.join(REPO_ROOT, ".pi", "batch-state.json");
-  BATCH_HISTORY_PATH = path.join(REPO_ROOT, ".pi", "batch-history.json");
-
-  const server = createServer();
-  const explicitPort = process.argv.slice(2).includes("--port");
-  const port = await findPort(server, opts.port, explicitPort);
-
-  console.log(`\n  Orchestrator Dashboard → http://localhost:${port}\n`);
-
-  // Broadcast state to all SSE clients on interval
-  const pollTimer = setInterval(broadcastState, POLL_INTERVAL);
-
-  // Also watch batch-state.json for immediate push on change
-  try {
-    const batchDir = path.dirname(BATCH_STATE_PATH);
-    if (fs.existsSync(batchDir)) {
-      let debounce = null;
-      fs.watch(batchDir, (eventType, filename) => {
-        if (filename === "batch-state.json") {
-          clearTimeout(debounce);
-          debounce = setTimeout(broadcastState, 200);
-        }
-      });
-    }
-  } catch {
-    // fs.watch not supported — polling is sufficient
-  }
-
-  // Auto-open browser
-  if (opts.open) {
-    setTimeout(() => openBrowser(`http://localhost:${port}`), 500);
-  }
-
-  // Graceful shutdown
-  function cleanup() {
-    clearInterval(pollTimer);
-    for (const client of sseClients) {
-      try { client.end(); } catch {}
-    }
-    server.close();
-    process.exit(0);
-  }
-
-  process.on("SIGINT", cleanup);
-  process.on("SIGTERM", cleanup);
-}
-
-main();
+#!/usr/bin/env node
+/**
+ * Orchestrator Web Dashboard — Local HTTP server with SSE live updates.
+ *
+ * Reads .pi/batch-state.json + STATUS.md files and streams state to the
+ * browser via Server-Sent Events. Zero external dependencies.
+ *
+ * Usage:
+ *   node dashboard/server.cjs [--port 8099] [--root /path/to/project]
+ */
+
+const http = require("http");
+const fs = require("fs");
+const path = require("path");
+const { exec } = require("child_process");
+// url module not needed — we parse with new URL() below
+
+// ─── Configuration ──────────────────────────────────────────────────────────
+
+const PUBLIC_DIR = path.join(__dirname, "public");
+const DEFAULT_PORT = 8100;
+const MAX_PORT_ATTEMPTS = 20;
+const POLL_INTERVAL = 2000; // ms between state checks
+
+// REPO_ROOT is resolved after parseArgs() — see initialization below.
+// In workspace mode, REPO_ROOT is the workspace root (passed via --root).
+// All dashboard state paths (batch-state, lane-state, conversation logs,
+// batch-history) live at <REPO_ROOT>/.pi/ — this is runtime/sidecar state
+// which does NOT follow the taskplane-pointer.json resolution chain.
+// The pointer directs config/agent lookups to a config repo, but the
+// dashboard only reads state files, so no pointer resolution is needed here.
+let REPO_ROOT;
+let BATCH_STATE_PATH;
+let BATCH_HISTORY_PATH;
+
+// ─── CLI Args ───────────────────────────────────────────────────────────────
+
+function parseArgs() {
+  const args = process.argv.slice(2);
+  const opts = { port: DEFAULT_PORT, open: true, root: "" };
+  for (let i = 0; i < args.length; i++) {
+    if (args[i] === "--port" && args[i + 1]) {
+      opts.port = parseInt(args[i + 1]) || DEFAULT_PORT;
+      i++;
+    } else if (args[i] === "--root" && args[i + 1]) {
+      opts.root = args[i + 1];
+      i++;
+    } else if (args[i] === "--no-open") {
+      opts.open = false;
+    } else if (args[i] === "--help" || args[i] === "-h") {
+      console.log(`
+Orchestrator Web Dashboard
+
+Usage:
+  node dashboard/server.cjs [options]
+
+Options:
+  --port <number>   Port to listen on (default: ${DEFAULT_PORT})
+  --root <path>     Project root directory (default: current directory)
+  --no-open         Don't auto-open browser
+  -h, --help        Show this help
+`);
+      process.exit(0);
+    }
+  }
+  return opts;
+}
+
+// ─── Data Loading (ported from orch-dashboard.cjs) ──────────────────────────
+
+function normalizeBatchStateIngress(state) {
+  if (!state || typeof state !== "object" || !Array.isArray(state.lanes)) {
+    return state;
+  }
+
+  for (const lane of state.lanes) {
+    if (!lane || typeof lane !== "object") continue;
+    // Legacy compatibility: older persisted states stored lane session IDs under
+    // `tmuxSessionName`. Normalize to `laneSessionId` at ingress and drop legacy key.
+    const laneSessionId = typeof lane.laneSessionId === "string"
+      ? lane.laneSessionId
+      : (typeof lane.tmuxSessionName === "string" ? lane.tmuxSessionName : undefined);
+    if (laneSessionId) {
+      lane.laneSessionId = laneSessionId;
+    }
+    if ("tmuxSessionName" in lane) {
+      delete lane.tmuxSessionName;
+    }
+  }
+
+  return state;
+}
+
+function loadBatchState() {
+  try {
+    const raw = fs.readFileSync(BATCH_STATE_PATH, "utf-8");
+    return normalizeBatchStateIngress(JSON.parse(raw));
+  } catch {
+    return null;
+  }
+}
+
+function resolveTaskFolder(task, state) {
+  if (!task || !task.taskFolder) return null;
+  const laneNum = task.laneNumber;
+  const lane = (state?.lanes || []).find((l) => l.laneNumber === laneNum);
+  if (!lane || !lane.worktreePath) return task.taskFolder;
+
+  // In workspace mode, the worktree is inside a specific repo, not the workspace root.
+  // The task folder path needs to be made relative to the repo root (parent of the worktree),
+  // not the workspace root. Detect this by finding the repo root from the worktree path.
+  const taskFolderAbs = path.resolve(task.taskFolder);
+  const worktreeAbs = path.resolve(lane.worktreePath);
+
+  // Try to find the repo root: walk up from the worktree path looking for which
+  // ancestor is a prefix of the task folder. The worktree is at <repoRoot>/.worktrees/<name>
+  // or a sibling, so the repo root is typically 2 levels up from a subdirectory worktree.
+  // Heuristic: find the longest common ancestor between taskFolder and worktree's repo root.
+  const repoRootAbs = path.resolve(REPO_ROOT);
+
+  // First try: relative to workspace root (works in repo mode where workspace = repo)
+  let rel = path.relative(repoRootAbs, taskFolderAbs);
+  if (!rel || rel.startsWith("..") || path.isAbsolute(rel)) return task.taskFolder;
+
+  // Check if joining with worktree produces a valid path
+  const candidate = path.join(worktreeAbs, rel);
+  try {
+    if (fs.existsSync(candidate)) return candidate;
+  } catch { /* fall through */ }
+
+  // Second try: the worktree is inside a repo subdirectory of the workspace root.
+  // Strip the repo prefix from the task folder path to get the repo-relative path.
+  // e.g., taskFolder = "workspace/platform-docs/task-mgmt/DOC-001/"
+  //        worktree  = "workspace/platform-docs/.worktrees/wt-1/"
+  //        repo-relative = "task-mgmt/DOC-001/"
+  // Find the repo by checking which workspace repo path is a prefix of the task folder.
+  const repoRoots = [];
+  try {
+    const stateMode = state.mode;
+    if (stateMode === "workspace" && state.repos) {
+      for (const r of state.repos) repoRoots.push(path.resolve(r.path));
+    }
+  } catch { /* no repo info in state */ }
+
+  // Also try inferring repo root from worktree path pattern:
+  // .worktrees/<name> → parent is repo root; sibling worktrees → shared parent
+  const worktreeParent = path.dirname(worktreeAbs);
+  const worktreeGrandparent = path.dirname(worktreeParent);
+  for (const possibleRepoRoot of [worktreeGrandparent, ...repoRoots]) {
+    const repoRel = path.relative(possibleRepoRoot, taskFolderAbs);
+    if (repoRel && !repoRel.startsWith("..") && !path.isAbsolute(repoRel)) {
+      const repoCandidate = path.join(worktreeAbs, repoRel);
+      try {
+        if (fs.existsSync(repoCandidate)) return repoCandidate;
+      } catch { continue; }
+    }
+  }
+
+  // Fallback: return original task folder (might work if not in worktree)
+  return task.taskFolder;
+}
+
+function parseStatusMd(taskFolder) {
+  const candidates = [taskFolder];
+  const taskId = path.basename(taskFolder);
+  const archiveBase = taskFolder.replace(/[/\\]tasks[/\\][^/\\]+$/, "/tasks/archive/" + taskId);
+  if (archiveBase !== taskFolder) candidates.push(archiveBase);
+
+  for (const folder of candidates) {
+    const statusPath = path.join(folder, "STATUS.md");
+    try {
+      const content = fs.readFileSync(statusPath, "utf-8");
+      const stepMatch = content.match(/\*\*Current Step:\*\*\s*(.+)/);
+      const statusMatch = content.match(/\*\*Status:\*\*\s*(.+)/);
+      const iterMatch = content.match(/\*\*Iteration:\*\*\s*(\d+)/);
+      const reviewMatch = content.match(/\*\*Review Counter:\*\*\s*(\d+)/);
+      const checked = (content.match(/- \[x\]/gi) || []).length;
+      const unchecked = (content.match(/- \[ \]/g) || []).length;
+      const total = checked + unchecked;
+      return {
+        currentStep: stepMatch ? stepMatch[1].trim() : "Unknown",
+        status: statusMatch ? statusMatch[1].trim() : "Unknown",
+        iteration: iterMatch ? parseInt(iterMatch[1]) : 0,
+        reviews: reviewMatch ? parseInt(reviewMatch[1]) : 0,
+        checked,
+        total,
+        progress: total > 0 ? Math.round((checked / total) * 100) : 0,
+      };
+    } catch {
+      continue;
+    }
+  }
+  return null;
+}
+
+function getActiveSessions() {
+  // Runtime V2: return active merger session names from the runtime registry
+  // so the dashboard merge pane can display live telemetry for running agents.
+  // Terminal statuses indicate the agent is no longer alive.
+  const TERMINAL_STATUSES = new Set(["exited", "killed", "crashed", "timed_out"]);
+  try {
+    const state = loadBatchState();
+    if (!state || !state.batchId) return [];
+    const registry = loadRuntimeRegistry(state.batchId);
+    if (!registry || !registry.agents) return [];
+    return Object.values(registry.agents)
+      .filter(a => a.role === "merger" && !TERMINAL_STATUSES.has(a.status))
+      .map(a => a.agentId);
+  } catch {
+    return [];
+  }
+}
+
+function checkDoneFile(taskFolder) {
+  const candidates = [taskFolder];
+  const taskId = path.basename(taskFolder);
+  const archiveBase = taskFolder.replace(/[/\\]tasks[/\\][^/\\]+$/, "/tasks/archive/" + taskId);
+  if (archiveBase !== taskFolder) candidates.push(archiveBase);
+  for (const folder of candidates) {
+    if (fs.existsSync(path.join(folder, ".DONE"))) return true;
+  }
+  return false;
+}
+
+/** Read lane state sidecar JSON files written by the task-runner. */
+function loadLaneStates() {
+  const piDir = path.join(REPO_ROOT, ".pi");
+  const states = {};
+  try {
+    const files = fs.readdirSync(piDir).filter(f => f.startsWith("lane-state-") && f.endsWith(".json"));
+    for (const file of files) {
+      try {
+        const raw = fs.readFileSync(path.join(piDir, file), "utf-8").trim();
+        if (!raw) continue;
+        const data = JSON.parse(raw);
+        if (data.prefix) states[data.prefix] = data;
+      } catch { continue; }
+    }
+  } catch { /* .pi dir may not exist */ }
+  return states;
+}
+
+// ─── Telemetry JSONL Tailing ────────────────────────────────────────────────
+
+/**
+ * Module-level tail state for incremental JSONL reading.
+ * Persists across poll ticks within this server process.
+ * Key: absolute file path → { offset, partial }
+ */
+const telemetryTailStates = new Map();
+
+/**
+ * Module-level accumulated telemetry per session prefix.
+ * Persists across poll ticks so incremental tail reads accumulate correctly.
+ * Key: session prefix → { inputTokens, outputTokens, ... }
+ */
+const telemetryAccumulators = new Map();
+
+/**
+ * Tracks which files are currently contributing to each prefix.
+ * Key: session prefix → Set of absolute file paths
+ * Used to detect file rotation: when files change, accumulator is reset.
+ */
+const telemetryPrefixFiles = new Map();
+
+/**
+ * Parse a telemetry JSONL filename to extract lane number and role.
+ * Pattern: {opId}-{batchId}-{repoId}[-{taskId}][-lane-{N}]-{role}.jsonl
+ * Roles: worker, reviewer, merger
+ * Returns { laneNumber: number|null, role: string, mergeNumber: number|null } or null if unparseable.
+ */
+function parseTelemetryFilename(filename) {
+  // Remove .jsonl extension
+  const base = filename.replace(/\.jsonl$/, "");
+  // Role is always the last segment
+  const lastDash = base.lastIndexOf("-");
+  if (lastDash < 0) return null;
+  const role = base.slice(lastDash + 1);
+  if (role !== "worker" && role !== "reviewer" && role !== "merger") return null;
+
+  // Extract lane number from -lane-{N}- pattern
+  const laneMatch = base.match(/-lane-(\d+)-/);
+  const laneNumber = laneMatch ? parseInt(laneMatch[1], 10) : null;
+
+  // Extract merge number from -merge-{N}- pattern (merge agents)
+  const mergeMatch = base.match(/-merge-(\d+)-/);
+  const mergeNumber = mergeMatch ? parseInt(mergeMatch[1], 10) : null;
+
+  return { laneNumber, role, mergeNumber };
+}
+
+/**
+ * Incrementally read new bytes from a JSONL file, parse events, and return them.
+ * Handles: file not yet created, empty reads, partial trailing lines, malformed JSON.
+ * @param {string} filePath - Absolute path to the JSONL file
+ * @returns {object[]} Array of parsed event objects from new data
+ */
+function tailJsonlFile(filePath) {
+  // Get or create tail state for this file
+  let tailState = telemetryTailStates.get(filePath);
+  if (!tailState) {
+    tailState = { offset: 0, partial: "" };
+    telemetryTailStates.set(filePath, tailState);
+  }
+
+  // Check file size
+  let fileSize;
+  try {
+    fileSize = fs.statSync(filePath).size;
+  } catch {
+    return []; // File doesn't exist yet
+  }
+
+  // Handle file truncation/recreation (offset beyond current size)
+  if (fileSize < tailState.offset) {
+    tailState.offset = 0;
+    tailState.partial = "";
+    tailState.wasReset = true; // Signal to caller that accumulator should be reset
+  }
+
+  if (fileSize <= tailState.offset) {
+    return []; // No new data
+  }
+
+  // Cap read size per tick to avoid ERR_STRING_TOO_LONG on large files.
+  // If there's more data remaining, the next SSE tick will pick up the rest.
+  const MAX_TAIL_BYTES = 10 * 1024 * 1024; // 10 MB per tick
+
+  // Skip-to-tail on fresh dashboard start with large files.
+  // The partial-line handling below already discards the first partial line.
+  if (tailState.offset === 0 && fileSize > MAX_TAIL_BYTES) {
+    tailState.offset = fileSize - MAX_TAIL_BYTES;
+  }
+
+  // Read new bytes from offset, capped to MAX_TAIL_BYTES
+  const bytesToRead = Math.min(fileSize - tailState.offset, MAX_TAIL_BYTES);
+  const buf = Buffer.alloc(bytesToRead);
+  let fd;
+  try {
+    fd = fs.openSync(filePath, "r");
+  } catch {
+    return []; // File became inaccessible
+  }
+  try {
+    fs.readSync(fd, buf, 0, bytesToRead, tailState.offset);
+  } catch {
+    fs.closeSync(fd);
+    return []; // Read error — try again next tick
+  }
+  fs.closeSync(fd);
+  tailState.offset += bytesToRead;
+
+  // Split into lines, preserving partial trailing line
+  const chunk = tailState.partial + buf.toString("utf-8");
+  const lines = chunk.split("\n");
+  tailState.partial = lines.pop() || "";
+
+  const events = [];
+  for (const line of lines) {
+    const trimmed = line.trim();
+    if (!trimmed) continue;
+    try {
+      const event = JSON.parse(trimmed);
+      if (event && event.type) events.push(event);
+    } catch {
+      // Malformed JSON — skip (concurrent write race, truncated line)
+    }
+  }
+  return events;
+}
+
+/**
+ * Load and accumulate telemetry from .pi/telemetry/*.jsonl files.
+ * Returns telemetry keyed by session prefix (e.g., "orch-lane-1").
+ *
+ * Uses batch-state lanes to map lane numbers → session prefixes.
+ * For standalone /task mode (no lane number in filename), data is keyed as "standalone".
+ *
+ * @param {object|null} batchState - The batch state from batch-state.json
+ * @returns {object} Map of sessionPrefix → accumulated telemetry
+ */
+
+// ── Runtime V2 Data Loaders (TP-107) ─────────────────────────────
+
+/**
+ * Load the Runtime V2 process registry for the current batch.
+ * Returns null if no registry exists (legacy batch).
+ */
+function loadRuntimeRegistry(batchId) {
+  if (!batchId) return null;
+  const registryPath = path.join(REPO_ROOT, ".pi", "runtime", batchId, "registry.json");
+  try {
+    if (!fs.existsSync(registryPath)) return null;
+    return JSON.parse(fs.readFileSync(registryPath, "utf-8"));
+  } catch {
+    return null;
+  }
+}
+
+/**
+ * Load Runtime V2 lane snapshots for the current batch.
+ * Returns a map of laneNumber → snapshot data.
+ */
+function loadRuntimeLaneSnapshots(batchId) {
+  if (!batchId) return {};
+  const lanesDir = path.join(REPO_ROOT, ".pi", "runtime", batchId, "lanes");
+  const snapshots = {};
+  try {
+    if (!fs.existsSync(lanesDir)) return snapshots;
+    const files = fs.readdirSync(lanesDir).filter(f => f.startsWith("lane-") && f.endsWith(".json"));
+    for (const file of files) {
+      try {
+        const data = JSON.parse(fs.readFileSync(path.join(lanesDir, file), "utf-8"));
+        if (data.laneNumber != null) snapshots[data.laneNumber] = data;
+      } catch { continue; }
+    }
+  } catch { /* dir missing */ }
+  return snapshots;
+}
+
+/**
+ * Load Runtime V2 merge agent snapshots for the current batch.
+ *
+ * Reads all `merge-N.json` files from `.pi/runtime/{batchId}/lanes/`.
+ * Returns a map of mergeNumber (string) → snapshot data.
+ *
+ * Follows the same pattern as {@link loadRuntimeLaneSnapshots}.
+ *
+ * @since TP-164
+ */
+function loadRuntimeMergeSnapshots(batchId) {
+  if (!batchId) return {};
+  const lanesDir = path.join(REPO_ROOT, ".pi", "runtime", batchId, "lanes");
+  const snapshots = {};
+  try {
+    if (!fs.existsSync(lanesDir)) return snapshots;
+    const files = fs.readdirSync(lanesDir).filter(f => f.startsWith("merge-") && f.endsWith(".json"));
+    for (const file of files) {
+      try {
+        const data = JSON.parse(fs.readFileSync(path.join(lanesDir, file), "utf-8"));
+        if (data.mergeNumber != null) snapshots[data.mergeNumber] = data;
+      } catch { continue; }
+    }
+  } catch { /* dir missing */ }
+  return snapshots;
+}
+
+/**
+ * Load Runtime V2 agent events for a specific agent.
+ * Returns the last N events from the agent's events.jsonl.
+ */
+function loadRuntimeAgentEvents(batchId, agentId, maxEvents) {
+  if (!batchId || !agentId) return [];
+  maxEvents = maxEvents || 200;
+  const eventsPath = path.join(REPO_ROOT, ".pi", "runtime", batchId, "agents", agentId, "events.jsonl");
+  try {
+    if (!fs.existsSync(eventsPath)) return [];
+    const raw = fs.readFileSync(eventsPath, "utf-8");
+    const lines = raw.split("\n").filter(l => l.trim());
+    const events = [];
+    const start = Math.max(0, lines.length - maxEvents);
+    for (let i = start; i < lines.length; i++) {
+      try { events.push(JSON.parse(lines[i])); } catch { continue; }
+    }
+    return events;
+  } catch {
+    return [];
+  }
+}
+
+/**
+ * Load mailbox message activity for the current batch.
+ *
+ * TP-093 hardening: event-authoritative model.
+ * Primary source: .pi/mailbox/{batchId}/events.jsonl (audit event stream).
+ * Fallback: directory scans (inbox/ack/outbox/outbox/processed) for
+ * compatibility when events.jsonl is absent.
+ *
+ * Includes:
+ * - Consumed replies (outbox/processed/) so they don't disappear after ack
+ * - Per-recipient broadcast delivery state from ack markers
+ * - Rate-limited events in the timeline
+ */
+function loadMailboxData(batchId) {
+  if (!batchId) return { messages: [], agentIds: [], auditEvents: [] };
+  const mbRoot = path.join(REPO_ROOT, ".pi", "mailbox", batchId);
+  if (!fs.existsSync(mbRoot)) return { messages: [], agentIds: [], auditEvents: [] };
+
+  // ── Primary: events.jsonl audit trail ──
+  const auditEvents = loadMailboxAuditEvents(mbRoot);
+
+  // ── Fallback: directory scan ──
+  const messages = [];
+  const agentIds = [];
+
+  try {
+    const dirs = fs.readdirSync(mbRoot, { withFileTypes: true })
+      .filter(d => d.isDirectory())
+      .map(d => d.name);
+
+    for (const agentDir of dirs) {
+      if (agentDir === "_broadcast") continue;
+      agentIds.push(agentDir);
+
+      // Scan inbox (pending), ack (delivered), outbox (active replies), outbox/processed (consumed replies)
+      for (const subdir of ["inbox", "ack", "outbox", "outbox/processed"]) {
+        const dir = path.join(mbRoot, agentDir, subdir);
+        if (!fs.existsSync(dir)) continue;
+        try {
+          const files = fs.readdirSync(dir).filter(f => f.endsWith(".msg.json"));
+          for (const file of files) {
+            try {
+              const msg = JSON.parse(fs.readFileSync(path.join(dir, file), "utf-8"));
+              let status;
+              if (subdir === "inbox") status = "pending";
+              else if (subdir === "ack") status = "delivered";
+              else if (subdir === "outbox") status = "reply";
+              else status = "reply-acked"; // outbox/processed
+              const isBroadcast = msg.to === "_broadcast";
+              messages.push({ ...msg, _status: status, _agentDir: agentDir, _isBroadcast: isBroadcast });
+            } catch { continue; }
+          }
+        } catch { continue; }
+      }
+    }
+
+    // _broadcast: per-recipient delivery state
+    const broadcastInbox = path.join(mbRoot, "_broadcast", "inbox");
+    const broadcastAck = path.join(mbRoot, "_broadcast", "ack");
+    for (const [dir, status] of [[broadcastInbox, "pending"], [broadcastAck, "delivered"]]) {
+      if (!fs.existsSync(dir)) continue;
+      try {
+        const files = fs.readdirSync(dir).filter(f => f.endsWith(".msg.json"));
+        for (const file of files) {
+          try {
+            const msg = JSON.parse(fs.readFileSync(path.join(dir, file), "utf-8"));
+            messages.push({ ...msg, _status: status, _agentDir: "_broadcast", _isBroadcast: true });
+          } catch { continue; }
+        }
+      } catch { continue; }
+    }
+  } catch { /* mailbox dir issues */ }
+
+  messages.sort((a, b) => (a.timestamp || 0) - (b.timestamp || 0));
+  return { messages, agentIds, auditEvents };
+}
+
+/**
+ * Load mailbox audit events from events.jsonl.
+ * Returns events sorted by timestamp. Includes: message_sent, message_delivered,
+ * message_replied, message_escalated, message_rate_limited.
+ */
+function loadMailboxAuditEvents(mbRoot) {
+  const eventsPath = path.join(mbRoot, "events.jsonl");
+  if (!fs.existsSync(eventsPath)) return [];
+  try {
+    const raw = fs.readFileSync(eventsPath, "utf-8");
+    const events = [];
+    for (const line of raw.split("\n")) {
+      if (!line.trim()) continue;
+      try { events.push(JSON.parse(line)); } catch { continue; }
+    }
+    return events;
+  } catch {
+    return [];
+  }
+}
+
+// ─────────────────────────────────────────────────────────────────────
+function loadTelemetryData(batchState) {
+  const telemetryDir = path.join(REPO_ROOT, ".pi", "telemetry");
+  const result = {};
+
+  // Build lane number → session prefix mapping from batch state
+  const laneToPrefix = {};
+  if (batchState && batchState.lanes) {
+    for (const lane of batchState.lanes) {
+      const laneSessionId = lane.laneSessionId;
+      if (lane.laneNumber != null && laneSessionId) {
+        laneToPrefix[lane.laneNumber] = laneSessionId;
+      }
+    }
+  }
+
+  // Scan telemetry directory for JSONL files
+  let files;
+  try {
+    files = fs.readdirSync(telemetryDir).filter(f => f.endsWith(".jsonl"));
+  } catch {
+    // .pi/telemetry/ may not exist (pre-RPC sessions) — degrade gracefully
+    return result;
+  }
+
+  // Track which files still exist for tail-state cleanup
+  const currentFiles = new Set();
+  // Track current file→prefix mapping to detect file rotation
+  const currentPrefixFiles = new Map(); // prefix → Set<filePath>
+
+  for (const file of files) {
+    const filePath = path.join(telemetryDir, file);
+    currentFiles.add(filePath);
+
+    // Parse filename to get lane number and role
+    const parsed = parseTelemetryFilename(file);
+    if (!parsed) continue;
+
+    // Determine the key (session prefix)
+    let prefix;
+    if (parsed.role === "merger") {
+      // Merge agent — derive prefix from lane naming so it matches the
+      // session ID used by the client (e.g. "orch-henrylach-merge-1").
+      // Lane sessions: "orch-{opId}-lane-{N}" → merge sessions: "orch-{opId}-merge-{N}".
+      const firstLanePrefix = Object.values(laneToPrefix)[0]; // e.g. "orch-henrylach-lane-1"
+      const opPrefix = firstLanePrefix?.replace(/-lane-\d+$/, ""); // "orch-henrylach"
+      if (parsed.mergeNumber != null && opPrefix) {
+        prefix = `${opPrefix}-merge-${parsed.mergeNumber}`;
+      } else if (parsed.mergeNumber != null) {
+        prefix = `orch-merge-${parsed.mergeNumber}`;
+      } else {
+        prefix = "orch-merge";
+      }
+    } else if (parsed.laneNumber != null && laneToPrefix[parsed.laneNumber]) {
+      prefix = laneToPrefix[parsed.laneNumber];
+    } else if (parsed.laneNumber != null) {
+      // Lane number found but no batch-state mapping — use heuristic
+      prefix = `orch-lane-${parsed.laneNumber}`;
+    } else {
+      // Standalone /task mode
+      prefix = "standalone";
+    }
+
+    // Track file→prefix mapping
+    if (!currentPrefixFiles.has(prefix)) currentPrefixFiles.set(prefix, new Set());
+    currentPrefixFiles.get(prefix).add(filePath);
+
+    // Check if file set for this prefix has changed (file rotation)
+    const prevFiles = telemetryPrefixFiles.get(prefix);
+    const isNewFile = !prevFiles || !prevFiles.has(filePath);
+
+    // Initialize persistent accumulator for this prefix if needed,
+    // or reset if files changed (new file appeared for same prefix)
+    if (!telemetryAccumulators.has(prefix) || (isNewFile && !telemetryTailStates.has(filePath))) {
+      const fresh = {
+        inputTokens: 0, outputTokens: 0, cacheReadTokens: 0,
+        cacheWriteTokens: 0, cost: 0, toolCalls: 0,
+        lastTool: "", currentTool: "", retries: 0, retryActive: false,
+        lastRetryError: "", compactions: 0, latestTotalTokens: 0,
+        contextPct: 0, startedAt: 0,
+      };
+      telemetryAccumulators.set(prefix, fresh);
+      // Also reset tail states for ALL files of this prefix to re-read from beginning
+      if (prevFiles) {
+        for (const pf of prevFiles) {
+          telemetryTailStates.delete(pf);
+        }
+      }
+    }
+
+    const acc = telemetryAccumulators.get(prefix);
+    result[prefix] = acc; // expose the persistent accumulator in the result
+
+    // Tail the file for new events
+    const events = tailJsonlFile(filePath);
+
+    // Check if file was truncated — reset accumulator
+    const ts = telemetryTailStates.get(filePath);
+    if (ts && ts.wasReset) {
+      acc.inputTokens = 0; acc.outputTokens = 0; acc.cacheReadTokens = 0;
+      acc.cacheWriteTokens = 0; acc.cost = 0; acc.toolCalls = 0;
+      acc.lastTool = ""; acc.currentTool = ""; acc.retries = 0; acc.retryActive = false;
+      acc.lastRetryError = ""; acc.compactions = 0; acc.latestTotalTokens = 0;
+      acc.contextPct = 0; acc.startedAt = 0;
+      ts.wasReset = false;
+    }
+    for (const event of events) {
+      switch (event.type) {
+        case "message_end": {
+          // A successful message_end means any prior retry resolved.
+          // Clear retryActive to prevent stale retry badges from persisting
+          // across batches or after transient API errors recover.
+          acc.retryActive = false;
+          const usage = event.message?.usage;
+          if (usage) {
+            acc.inputTokens += usage.input || 0;
+            acc.outputTokens += usage.output || 0;
+            acc.cacheReadTokens += usage.cacheRead || 0;
+            acc.cacheWriteTokens += usage.cacheWrite || 0;
+            if (usage.cost) {
+              acc.cost += typeof usage.cost === "object"
+                ? (usage.cost.total || 0)
+                : (typeof usage.cost === "number" ? usage.cost : 0);
+            }
+            // Include cacheRead: totalTokens from pi excludes cache reads,
+            // but cached tokens still consume context window capacity.
+            const rawTotal = usage.totalTokens
+              || ((usage.input || 0) + (usage.output || 0));
+            const totalTokens = rawTotal + (usage.cacheRead || 0);
+            if (totalTokens > acc.latestTotalTokens) {
+              acc.latestTotalTokens = totalTokens;
+            }
+          }
+          break;
+        }
+        case "tool_execution_start": {
+          acc.toolCalls++;
+          const toolDesc = event.toolName || "unknown";
+          let argPreview = "";
+          if (event.args) {
+            if (typeof event.args === "string") {
+              argPreview = event.args.slice(0, 80);
+            } else if (typeof event.args === "object") {
+              const firstVal = Object.values(event.args)[0];
+              if (typeof firstVal === "string") {
+                argPreview = firstVal.slice(0, 80);
+              }
+            }
+          }
+          const toolLabel = argPreview ? `${toolDesc} ${argPreview}` : toolDesc;
+          acc.lastTool = toolLabel;
+          acc.currentTool = toolLabel;
+          break;
+        }
+        case "tool_execution_end": {
+          acc.currentTool = "";
+          break;
+        }
+        case "agent_start": {
+          if (event.ts && !acc.startedAt) {
+            acc.startedAt = event.ts;
+          }
+          break;
+        }
+        case "response": {
+          // Extract context usage from get_session_stats responses
+          const ctxUsage = event.data?.contextUsage;
+          if (ctxUsage) {
+            // Support both percent (current) and percentUsed (legacy pi versions)
+            const pct = typeof ctxUsage.percent === "number" ? ctxUsage.percent
+              : typeof ctxUsage.percentUsed === "number" ? ctxUsage.percentUsed
+              : null;
+            if (pct !== null) acc.contextPct = pct;
+          }
+          break;
+        }
+        case "auto_retry_start": {
+          acc.retries++;
+          acc.retryActive = true;
+          acc.lastRetryError = event.errorMessage || event.error || "unknown";
+          break;
+        }
+        case "auto_retry_end": {
+          acc.retryActive = false;
+          break;
+        }
+        case "auto_compaction_start": {
+          acc.compactions++;
+          break;
+        }
+      }
+    }
+  }
+
+  // Clean up tail states for files that no longer exist
+  for (const [filePath] of telemetryTailStates) {
+    if (filePath.startsWith(telemetryDir) && !currentFiles.has(filePath)) {
+      telemetryTailStates.delete(filePath);
+    }
+  }
+
+  // Update prefix→files tracking for next call
+  // Clean up accumulators and tracking for prefixes that have no remaining files
+  const activePrefixes = new Set(Object.keys(result));
+  for (const [prefix] of telemetryAccumulators) {
+    if (!activePrefixes.has(prefix)) {
+      telemetryAccumulators.delete(prefix);
+      telemetryPrefixFiles.delete(prefix);
+    }
+  }
+  // Store current file mappings for next call's rotation detection
+  for (const [prefix, fileSet] of currentPrefixFiles) {
+    telemetryPrefixFiles.set(prefix, fileSet);
+  }
+
+  return result;
+}
+
+// ─── Supervisor Data Loading ────────────────────────────────────────────────
+
+/**
+ * Module-level tail state for supervisor JSONL files (actions.jsonl, events.jsonl).
+ * Reuses the same incremental tailing pattern as telemetry.
+ * Key: absolute file path → { offset, partial, entries }
+ */
+const supervisorTailStates = {
+  actions: { offset: 0, partial: "", entries: [] },
+  events: { offset: 0, partial: "", entries: [] },
+  conversation: { offset: 0, partial: "", entries: [] },
+};
+
+/**
+ * The last known batchId — used to detect batch changes and reset accumulators.
+ */
+let supervisorLastBatchId = "";
+
+/**
+ * Incrementally tail a JSONL file, accumulating parsed entries.
+ * Filters entries by batchId when provided.
+ *
+ * @param {string} filePath - Absolute path to the JSONL file
+ * @param {object} tailState - Mutable tail state { offset, partial, entries }
+ * @param {string} batchId - Batch ID to filter by (empty = no filter)
+ * @returns {object[]} The accumulated entries array (same reference as tailState.entries)
+ */
+function tailSupervisorJsonl(filePath, tailState, batchId) {
+  // Check file size
+  let fileSize;
+  try {
+    fileSize = fs.statSync(filePath).size;
+  } catch {
+    return tailState.entries; // File doesn't exist yet — return accumulated
+  }
+
+  // Handle file truncation/recreation
+  if (fileSize < tailState.offset) {
+    tailState.offset = 0;
+    tailState.partial = "";
+    tailState.entries = [];
+  }
+
+  if (fileSize <= tailState.offset) {
+    return tailState.entries; // No new data
+  }
+
+  // Read new bytes from offset
+  const bytesToRead = fileSize - tailState.offset;
+  const buf = Buffer.alloc(bytesToRead);
+  let fd;
+  try {
+    fd = fs.openSync(filePath, "r");
+  } catch {
+    return tailState.entries;
+  }
+  try {
+    fs.readSync(fd, buf, 0, bytesToRead, tailState.offset);
+  } catch {
+    fs.closeSync(fd);
+    return tailState.entries;
+  }
+  fs.closeSync(fd);
+  tailState.offset = fileSize;
+
+  // Split into lines, preserving partial trailing line
+  const chunk = tailState.partial + buf.toString("utf-8");
+  const lines = chunk.split("\n");
+  tailState.partial = lines.pop() || "";
+
+  for (const line of lines) {
+    const trimmed = line.trim();
+    if (!trimmed) continue;
+    try {
+      const entry = JSON.parse(trimmed);
+      // Filter by batchId if provided
+      if (batchId && entry.batchId && entry.batchId !== batchId) continue;
+      tailState.entries.push(entry);
+    } catch {
+      // Malformed JSON — skip
+    }
+  }
+
+  // Cap accumulated entries to prevent unbounded growth (keep last 500)
+  if (tailState.entries.length > 500) {
+    tailState.entries = tailState.entries.slice(-500);
+  }
+
+  return tailState.entries;
+}
+
+/**
+ * Read supervisor autonomy level from project config.
+ *
+ * Checks `.pi/taskplane-config.json` for `orchestrator.supervisor.autonomy`.
+ * Falls back to "supervised" (the default) if config is missing or malformed.
+ * This is needed because the lockfile does not contain the autonomy level.
+ *
+ * @returns {string} Autonomy level: "interactive" | "supervised" | "autonomous"
+ */
+function loadSupervisorAutonomy() {
+  try {
+    const configPath = path.join(REPO_ROOT, ".pi", "taskplane-config.json");
+    const raw = fs.readFileSync(configPath, "utf-8");
+    const config = JSON.parse(raw);
+    const autonomy = config?.orchestrator?.supervisor?.autonomy;
+    if (autonomy === "interactive" || autonomy === "supervised" || autonomy === "autonomous") {
+      return autonomy;
+    }
+  } catch {
+    // Config missing or malformed — use default
+  }
+  return "supervised"; // Default per DEFAULT_SUPERVISOR_CONFIG
+}
+
+/**
+ * Load supervisor data for the dashboard.
+ *
+ * Reads (all from .pi/supervisor/):
+ * - lock.json: supervisor active/stale status, heartbeat, autonomy (from config)
+ * - actions.jsonl: recovery action audit trail (batch-scoped, incremental)
+ * - events.jsonl: engine + tier 0 events (batch-scoped, incremental)
+ * - conversation.jsonl: operator ↔ supervisor interaction log (spec §9.1)
+ * - summary.md: human-readable batch summary (generated on completion)
+ *
+ * Returns null when no supervisor files exist (pre-supervisor batches).
+ *
+ * @param {object|null} batchState - The batch state from batch-state.json
+ * @returns {object|null} Supervisor data object or null
+ */
+function loadSupervisorData(batchState) {
+  const supervisorDir = path.join(REPO_ROOT, ".pi", "supervisor");
+  const batchId = batchState ? (batchState.batchId || "") : "";
+
+  // Detect batch change — reset tail state accumulators
+  if (batchId && batchId !== supervisorLastBatchId) {
+    supervisorLastBatchId = batchId;
+    supervisorTailStates.actions = { offset: 0, partial: "", entries: [] };
+    supervisorTailStates.events = { offset: 0, partial: "", entries: [] };
+    supervisorTailStates.conversation = { offset: 0, partial: "", entries: [] };
+  }
+
+  // ── Lockfile: supervisor status ──
+  // The lockfile contains pid, sessionId, batchId, startedAt, heartbeat.
+  // It does NOT contain autonomy — that comes from project config.
+  let lock = null;
+  try {
+    const lockPath = path.join(supervisorDir, "lock.json");
+    const raw = fs.readFileSync(lockPath, "utf-8");
+    const parsed = JSON.parse(raw);
+    if (parsed && parsed.pid && parsed.sessionId) {
+      // Determine if lock is stale (heartbeat older than 90s)
+      const heartbeatAge = parsed.heartbeat
+        ? Date.now() - new Date(parsed.heartbeat).getTime()
+        : Infinity;
+      const isStale = heartbeatAge > 90_000;
+
+      lock = {
+        active: !isStale,
+        pid: parsed.pid,
+        sessionId: parsed.sessionId,
+        batchId: parsed.batchId || "",
+        startedAt: parsed.startedAt || "",
+        heartbeat: parsed.heartbeat || "",
+        // Autonomy is NOT in the lockfile — derive from project config
+        autonomy: loadSupervisorAutonomy(),
+      };
+    }
+  } catch {
+    // No lockfile or malformed — supervisor is inactive
+  }
+
+  // ── Actions JSONL: recovery audit trail (batch-scoped, incremental) ──
+  const actionsPath = path.join(supervisorDir, "actions.jsonl");
+  const actions = tailSupervisorJsonl(actionsPath, supervisorTailStates.actions, batchId);
+
+  // ── Events JSONL: engine events (batch-scoped, incremental) ──
+  const eventsPath = path.join(supervisorDir, "events.jsonl");
+  const events = tailSupervisorJsonl(eventsPath, supervisorTailStates.events, batchId);
+
+  // ── Conversation JSONL: operator interaction log (spec §9.1) ──
+  // The supervisor writes operator↔supervisor messages to conversation.jsonl.
+  // Not yet implemented in all supervisor versions — degrade gracefully.
+  const conversationPath = path.join(supervisorDir, "conversation.jsonl");
+  const conversation = tailSupervisorJsonl(conversationPath, supervisorTailStates.conversation, batchId);
+
+  // ── Summary: human-readable batch summary (generated on completion) ──
+  // Per spec §9.1, the supervisor writes .pi/supervisor/summary.md when the
+  // batch completes or is abandoned. Read the file if it exists.
+  let summary = null;
+  try {
+    const summaryPath = path.join(supervisorDir, "summary.md");
+    summary = fs.readFileSync(summaryPath, "utf-8");
+  } catch {
+    // No summary yet — batch may still be running, or pre-supervisor batch
+  }
+
+  // If nothing exists at all, return null (pre-supervisor batch)
+  if (!lock && actions.length === 0 && events.length === 0 && conversation.length === 0 && !summary) {
+    // Check if the supervisor directory even exists
+    try {
+      fs.statSync(supervisorDir);
+    } catch {
+      return null; // No supervisor dir → pre-supervisor batch
+    }
+  }
+
+  return {
+    lock,
+    actions,
+    events,
+    conversation,
+    summary,
+  };
+}
+
+/**
+ * Compute batch total cost from lane states (primary) and telemetry (supplementary).
+ * Lane states are authoritative — telemetry provides additional data only for lanes
+ * that have no lane-state entry (e.g., very early in session startup).
+ */
+function computeBatchTotalCost(laneStates, telemetry) {
+  let totalCost = 0;
+  const coveredPrefixes = new Set();
+
+  // Primary: sum cost from lane states (worker + reviewer)
+  for (const [prefix, ls] of Object.entries(laneStates)) {
+    if (ls.workerCostUsd) {
+      totalCost += ls.workerCostUsd;
+      coveredPrefixes.add(prefix);
+    }
+    if (ls.reviewerCostUsd) {
+      totalCost += ls.reviewerCostUsd;
+    }
+  }
+
+  // Supplementary: add cost from telemetry for uncovered lanes only
+  for (const [prefix, tel] of Object.entries(telemetry)) {
+    if (!coveredPrefixes.has(prefix) && tel.cost > 0) {
+      totalCost += tel.cost;
+    }
+  }
+
+  return totalCost;
+}
+
+function synthesizeLaneStateFromSnapshot(key, snap, fallbackBatchId) {
+  const w = snap.worker || {};
+  const r = snap.reviewer || null;
+  const statusMap = { running: "running", spawning: "running", exited: "done", crashed: "error", killed: "error", timed_out: "error", wrapping_up: "running" };
+  const reviewerStatusMap = { running: "running", spawning: "running", wrapping_up: "running", exited: "done", crashed: "done", killed: "done", timed_out: "done" };
+
+  return {
+    prefix: key,
+    taskId: snap.taskId || null,
+    phase: snap.status === "running" ? "worker-active" : snap.status === "complete" ? "complete" : "idle",
+    workerStatus: statusMap[w.status] || w.status || "idle",
+    workerElapsed: w.elapsedMs || 0,
+    workerContextPct: w.contextPct || 0,
+    workerLastTool: w.lastTool || "",
+    workerToolCount: w.toolCalls || 0,
+    workerInputTokens: w.inputTokens || 0,
+    workerOutputTokens: w.outputTokens || 0,
+    workerCacheReadTokens: w.cacheReadTokens || 0,
+    workerCacheWriteTokens: w.cacheWriteTokens || 0,
+    workerCostUsd: w.costUsd || 0,
+    reviewerStatus: r ? (reviewerStatusMap[r.status] || r.status || "running") : "idle",
+    reviewerElapsed: r?.elapsedMs || 0,
+    reviewerContextPct: r?.contextPct || 0,
+    reviewerLastTool: r?.lastTool || "",
+    reviewerToolCount: r?.toolCalls || 0,
+    reviewerCostUsd: r?.costUsd || 0,
+    reviewerInputTokens: r?.inputTokens || 0,
+    reviewerOutputTokens: r?.outputTokens || 0,
+    reviewerCacheReadTokens: r?.cacheReadTokens || 0,
+    reviewerCacheWriteTokens: r?.cacheWriteTokens || 0,
+    reviewerType: r?.reviewType || "",
+    reviewerStep: r?.reviewStep || 0,
+    batchId: snap.batchId || fallbackBatchId,
+    timestamp: snap.updatedAt || Date.now(),
+  };
+}
+
+/** Build full dashboard state object for the frontend. */
+function buildDashboardState() {
+  const state = loadBatchState();
+  const sessions = getActiveSessions();
+  const rawLaneStates = loadLaneStates();
+  // Filter stale lane states from previous batches.
+  // Lane state files persist across batches (same filename), so without
+  // filtering the dashboard shows telemetry from old runs.
+  const currentBatchId = state?.batchId || null;
+  const laneStates = {};
+  for (const [key, ls] of Object.entries(rawLaneStates)) {
+    if (!currentBatchId || !ls.batchId || ls.batchId === currentBatchId) {
+      laneStates[key] = ls;
+    }
+  }
+  const telemetry = loadTelemetryData(state);
+  const batchTotalCost = computeBatchTotalCost(laneStates, telemetry);
+  const supervisor = loadSupervisorData(state);
+
+  if (!state) {
+    return {
+      batch: null,
+      sessions,
+      tmuxSessions: sessions, // Legacy compatibility field for older dashboard clients
+      laneStates: {},
+      telemetry: {},
+      batchTotalCost: 0,
+      supervisor: null,
+      timestamp: Date.now(),
+    };
+  }
+
+  const tasks = (state.tasks || []).map((task) => {
+    const effectiveFolder = resolveTaskFolder(task, state);
+    let statusData = null;
+    if (effectiveFolder) {
+      statusData = parseStatusMd(effectiveFolder);
+    }
+    if (!task.doneFileFound && effectiveFolder) {
+      task.doneFileFound = checkDoneFile(effectiveFolder);
+    }
+    return { ...task, statusData };
+  });
+
+  // TP-107: Load Runtime V2 data when available
+  const runtimeRegistry = loadRuntimeRegistry(state.batchId);
+  const runtimeLaneSnapshots = loadRuntimeLaneSnapshots(state.batchId);
+  const mailboxData = loadMailboxData(state.batchId);
+
+  // TP-164: Load merge agent snapshots for live dashboard telemetry.
+  const runtimeMergeSnapshots = loadRuntimeMergeSnapshots(state.batchId);
+
+  // TP-164: Inject merge snapshot telemetry into the telemetry map so
+  // `telemetry[sessionName]` resolves for the merge pane.
+  // Snapshots are the PRIMARY source for merge telemetry — merge agents don't
+  // write to .pi/telemetry/*.jsonl (the JSONL path is for lane workers only).
+  // Inject snapshot data when no JSONL-backed entry exists for this session.
+  for (const snap of Object.values(runtimeMergeSnapshots)) {
+    const key = snap.sessionName;
+    if (!key) continue;
+    const agent = snap.agent;
+    if (!agent) continue;
+    // Only inject if there is no existing JSONL-backed telemetry entry.
+    // Merge agents don't emit to .pi/telemetry/, so existing will always be
+    // absent unless something else wrote to it — in that case defer to it.
+    const existing = telemetry[key];
+    if (!existing) {
+      telemetry[key] = {
+        inputTokens: agent.inputTokens || 0,
+        outputTokens: agent.outputTokens || 0,
+        cacheReadTokens: agent.cacheReadTokens || 0,
+        cacheWriteTokens: agent.cacheWriteTokens || 0,
+        cost: agent.costUsd || 0,
+        toolCalls: agent.toolCalls || 0,
+        lastTool: agent.lastTool || "",
+        currentTool: snap.status === "running" ? (agent.lastTool || "") : "",
+        contextPct: agent.contextPct || 0,
+        // startedAt is not in the snapshot; compute from elapsed if possible.
+        startedAt: agent.elapsedMs > 0 ? snap.updatedAt - agent.elapsedMs : snap.updatedAt,
+        retries: 0,
+        retryActive: false,
+        lastRetryError: "",
+        compactions: 0,
+        latestTotalTokens: (agent.inputTokens || 0) + (agent.outputTokens || 0),
+        _updatedAt: snap.updatedAt,
+        _source: "merge-snapshot",
+        // TP-178: Include waveIndex for precise wave-telemetry association (#498)
+        waveIndex: snap.waveIndex != null ? snap.waveIndex : undefined,
+      };
+    }
+  }
+
+  // TP-115: Synthesize laneStates from V2 snapshots so the dashboard
+  // pipeline works without legacy lane-state-*.json sidecar files.
+  // V2 snapshots are authoritative when present.
+  if (Object.keys(runtimeLaneSnapshots).length > 0) {
+    for (const [laneNum, snap] of Object.entries(runtimeLaneSnapshots)) {
+      // Find the matching lane record to get the session name key
+      const laneRec = (state.lanes || []).find(l => l.laneNumber === Number(laneNum));
+      const key = laneRec ? (laneRec.laneSessionId) : `lane-${laneNum}`;
+      if (!laneStates[key] || (snap.updatedAt && snap.updatedAt > (laneStates[key].timestamp || 0))) {
+        laneStates[key] = synthesizeLaneStateFromSnapshot(key, snap, state.batchId);
+      }
+    }
+  }
+
+  return {
+    laneStates,
+    telemetry,
+    batchTotalCost,
+    supervisor,
+    // Runtime V2 data (null/empty for legacy batches)
+    runtimeRegistry,
+    runtimeLaneSnapshots,
+    // TP-164: Merge agent snapshots for live dashboard telemetry.
+    runtimeMergeSnapshots,
+    mailbox: mailboxData,
+    batch: {
+      batchId: state.batchId,
+      phase: state.phase,
+      startedAt: state.startedAt,
+      updatedAt: state.updatedAt,
+      currentWaveIndex: state.currentWaveIndex || 0,
+      totalWaves: state.totalWaves || (state.wavePlan ? state.wavePlan.length : 0),
+      wavePlan: state.wavePlan || [],
+      // Lanes already include repoId (string|undefined) from PersistedLaneRecord (v2).
+      lanes: state.lanes || [],
+      // Tasks already include repoId, resolvedRepoId (string|undefined) from PersistedTaskRecord (v2).
+      tasks,
+      mergeResults: state.mergeResults || [],
+      errors: state.errors || [],
+      lastError: state.lastError || null,
+      // Workspace mode: "repo" (default/v1) or "workspace" (v2 multi-repo).
+      // Additive field — absent in v1 state files, frontend must default to "repo".
+      mode: state.mode || "repo",
+      // TP-148: Segment records for wave display context (v4+).
+      // Each record has taskId, segmentId, repoId, status.
+      segments: state.segments || [],
+    },
+    sessions,
+    tmuxSessions: sessions, // Legacy compatibility field for older dashboard clients
+    timestamp: Date.now(),
+  };
+}
+
+// ─── Static File Serving ────────────────────────────────────────────────────
+
+const MIME_TYPES = {
+  ".html": "text/html; charset=utf-8",
+  ".css": "text/css; charset=utf-8",
+  ".js": "application/javascript; charset=utf-8",
+  ".json": "application/json; charset=utf-8",
+  ".svg": "image/svg+xml",
+  ".png": "image/png",
+  ".ico": "image/x-icon",
+};
+
+function serveStatic(req, res) {
+  let filePath = new URL(req.url, "http://localhost").pathname;
+  if (filePath === "/") filePath = "/index.html";
+
+  const fullPath = path.join(PUBLIC_DIR, filePath);
+  // Prevent directory traversal
+  if (!fullPath.startsWith(PUBLIC_DIR)) {
+    res.writeHead(403);
+    res.end("Forbidden");
+    return;
+  }
+
+  const ext = path.extname(fullPath);
+  const contentType = MIME_TYPES[ext] || "application/octet-stream";
+
+  try {
+    const content = fs.readFileSync(fullPath);
+    res.writeHead(200, {
+      "Content-Type": contentType,
+      "Cache-Control": "no-cache, no-store, must-revalidate",
+    });
+    res.end(content);
+  } catch {
+    res.writeHead(404, { "Content-Type": "text/plain" });
+    res.end("Not Found");
+  }
+}
+
+// ─── SSE Stream ─────────────────────────────────────────────────────────────
+
+const sseClients = new Set();
+
+// ─── Conversation JSONL ─────────────────────────────────────────────────
+
+function serveConversation(req, res, prefix) {
+  if (!/^[\w-]+$/.test(prefix)) {
+    res.writeHead(400, { "Content-Type": "text/plain" });
+    res.end("Invalid prefix");
+    return;
+  }
+
+  const filePath = path.join(REPO_ROOT, ".pi", `worker-conversation-${prefix}.jsonl`);
+  try {
+    const content = fs.readFileSync(filePath, "utf-8");
+    res.writeHead(200, {
+      "Content-Type": "application/x-ndjson",
+      "Access-Control-Allow-Origin": "*",
+    });
+    res.end(content);
+  } catch {
+    res.writeHead(200, {
+      "Content-Type": "application/x-ndjson",
+      "Access-Control-Allow-Origin": "*",
+    });
+    res.end(""); // empty — no conversation yet
+  }
+}
+
+// ─── Dashboard SSE ──────────────────────────────────────────────────────
+
+function handleSSE(req, res) {
+  res.writeHead(200, {
+    "Content-Type": "text/event-stream",
+    "Cache-Control": "no-cache",
+    Connection: "keep-alive",
+    "Access-Control-Allow-Origin": "*",
+  });
+
+  // Send initial state immediately
+  const state = buildDashboardState();
+  res.write(`data: ${JSON.stringify(state)}\n\n`);
+
+  sseClients.add(res);
+  req.on("close", () => sseClients.delete(res));
+}
+
+function broadcastState() {
+  if (sseClients.size === 0) return;
+  const state = buildDashboardState();
+  const payload = `data: ${JSON.stringify(state)}\n\n`;
+  for (const client of sseClients) {
+    try {
+      client.write(payload);
+    } catch {
+      sseClients.delete(client);
+    }
+  }
+}
+
+// ─── Batch History API ──────────────────────────────────────────────────────
+
+// BATCH_HISTORY_PATH is initialized in main() alongside REPO_ROOT.
+
+function loadHistory() {
+  try {
+    if (!fs.existsSync(BATCH_HISTORY_PATH)) return [];
+    const raw = fs.readFileSync(BATCH_HISTORY_PATH, "utf-8");
+    return JSON.parse(raw);
+  } catch {
+    return [];
+  }
+}
+
+/** GET /api/history — return list of batch summaries (compact: no per-task detail). */
+function serveHistory(req, res) {
+  const history = loadHistory();
+  // Return compact list for the dropdown (no per-task details)
+  const compact = history.map(h => ({
+    batchId: h.batchId,
+    status: h.status,
+    startedAt: h.startedAt,
+    endedAt: h.endedAt,
+    durationMs: h.durationMs,
+    totalWaves: h.totalWaves,
+    totalTasks: h.totalTasks,
+    succeededTasks: h.succeededTasks,
+    failedTasks: h.failedTasks,
+    tokens: h.tokens,
+  }));
+  res.writeHead(200, {
+    "Content-Type": "application/json",
+    "Access-Control-Allow-Origin": "*",
+  });
+  res.end(JSON.stringify(compact));
+}
+
+/** GET /api/history/:batchId — return full detail for one batch. */
+function serveHistoryEntry(req, res, batchId) {
+  const history = loadHistory();
+  const entry = history.find(h => h.batchId === batchId);
+  if (!entry) {
+    res.writeHead(404, { "Content-Type": "application/json" });
+    res.end(JSON.stringify({ error: "Batch not found" }));
+    return;
+  }
+  res.writeHead(200, {
+    "Content-Type": "application/json",
+    "Access-Control-Allow-Origin": "*",
+  });
+  res.end(JSON.stringify(entry));
+}
+
+/** GET /api/status-md/:taskId — return raw STATUS.md content for a task. */
+function serveStatusMd(req, res, taskId) {
+  if (!/^[\w-]+$/.test(taskId)) {
+    res.writeHead(400, { "Content-Type": "text/plain" });
+    res.end("Invalid task ID");
+    return;
+  }
+
+  const state = loadBatchState();
+  if (!state) {
+    res.writeHead(404, { "Content-Type": "application/json" });
+    res.end(JSON.stringify({ error: "No batch state" }));
+    return;
+  }
+
+  const task = (state.tasks || []).find(t => t.taskId === taskId);
+  if (!task) {
+    res.writeHead(404, { "Content-Type": "application/json" });
+    res.end(JSON.stringify({ error: "Task not found" }));
+    return;
+  }
+
+  const effectiveFolder = resolveTaskFolder(task, state);
+  if (!effectiveFolder) {
+    res.writeHead(404, { "Content-Type": "application/json" });
+    res.end(JSON.stringify({ error: "Cannot resolve task folder" }));
+    return;
+  }
+
+  // Try effective folder, then archive
+  const candidates = [effectiveFolder];
+  const archiveBase = effectiveFolder.replace(/[/\\]tasks[/\\][^/\\]+$/, "/tasks/archive/" + taskId);
+  if (archiveBase !== effectiveFolder) candidates.push(archiveBase);
+
+  for (const folder of candidates) {
+    const statusPath = path.join(folder, "STATUS.md");
+    try {
+      const content = fs.readFileSync(statusPath, "utf-8");
+      res.writeHead(200, {
+        "Content-Type": "text/plain; charset=utf-8",
+        "Access-Control-Allow-Origin": "*",
+      });
+      res.end(content);
+      return;
+    } catch { continue; }
+  }
+
+  res.writeHead(404, { "Content-Type": "application/json" });
+  res.end(JSON.stringify({ error: "STATUS.md not found" }));
+}
+
+// ─── Dashboard Preferences ──────────────────────────────────────────────────
+
+function getPreferencesPath() {
+  return path.join(REPO_ROOT, ".pi", "dashboard-preferences.json");
+}
+
+function handleGetPreferences(req, res) {
+  const prefsPath = getPreferencesPath();
+  let prefs = { theme: "dark" };
+  try {
+    if (fs.existsSync(prefsPath)) {
+      prefs = JSON.parse(fs.readFileSync(prefsPath, "utf8"));
+    }
+  } catch { /* use defaults */ }
+  res.writeHead(200, {
+    "Content-Type": "application/json",
+    "Access-Control-Allow-Origin": "*",
+  });
+  res.end(JSON.stringify(prefs));
+}
+
+function handlePostPreferences(req, res) {
+  let body = "";
+  req.on("data", (chunk) => { body += chunk; });
+  req.on("end", () => {
+    try {
+      const incoming = JSON.parse(body);
+      const prefsPath = getPreferencesPath();
+      let existing = {};
+      try {
+        if (fs.existsSync(prefsPath)) {
+          existing = JSON.parse(fs.readFileSync(prefsPath, "utf8"));
+        }
+      } catch { /* start fresh */ }
+      const merged = { ...existing, ...incoming };
+      const dir = path.dirname(prefsPath);
+      if (!fs.existsSync(dir)) fs.mkdirSync(dir, { recursive: true });
+      fs.writeFileSync(prefsPath, JSON.stringify(merged, null, 2) + "\n");
+      res.writeHead(200, {
+        "Content-Type": "application/json",
+        "Access-Control-Allow-Origin": "*",
+      });
+      res.end(JSON.stringify(merged));
+    } catch (err) {
+      res.writeHead(400, {
+        "Content-Type": "application/json",
+        "Access-Control-Allow-Origin": "*",
+      });
+      res.end(JSON.stringify({ error: "Invalid JSON" }));
+    }
+  });
+}
+
+// ─── HTTP Server ────────────────────────────────────────────────────────────
+
+function createServer() {
+  const server = http.createServer((req, res) => {
+    const pathname = new URL(req.url, "http://localhost").pathname;
+
+    if (pathname === "/api/stream" && req.method === "GET") {
+      handleSSE(req, res);
+    } else if (pathname.startsWith("/api/conversation/") && req.method === "GET") {
+      const prefix = pathname.slice("/api/conversation/".length);
+      serveConversation(req, res, prefix);
+    } else if (pathname.startsWith("/api/agent-events/") && req.method === "GET") {
+      // TP-107: Serve Runtime V2 agent events (hardened)
+      const agentId = decodeURIComponent(pathname.slice("/api/agent-events/".length));
+      // Strict validation: same pattern as /api/conversation/:prefix
+      if (!/^[\w-]+$/.test(agentId)) {
+        res.writeHead(400, { "Content-Type": "text/plain" });
+        res.end("Invalid agent ID");
+        return;
+      }
+      const batchState = loadBatchState();
+      // Path containment: verify resolved path stays inside runtime dir
+      if (batchState?.batchId) {
+        const runtimeBase = path.join(REPO_ROOT, ".pi", "runtime", batchState.batchId, "agents");
+        const resolvedAgent = path.resolve(runtimeBase, agentId);
+        if (!resolvedAgent.startsWith(path.resolve(runtimeBase))) {
+          res.writeHead(403, { "Content-Type": "text/plain" });
+          res.end("Forbidden");
+          return;
+        }
+      }
+      // Optional: ?sinceTs= to return only events after a timestamp
+      const reqUrl = new URL(req.url, "http://localhost");
+      const sinceTs = parseInt(reqUrl.searchParams.get("sinceTs") || "0", 10);
+      let events = loadRuntimeAgentEvents(batchState?.batchId, agentId, 300);
+      if (sinceTs > 0) {
+        events = events.filter(e => (e.ts || 0) > sinceTs);
+      }
+      res.writeHead(200, { "Content-Type": "application/json", "Access-Control-Allow-Origin": "*" });
+      res.end(JSON.stringify(events));
+    } else if (pathname === "/api/state" && req.method === "GET") {
+      const state = buildDashboardState();
+      res.writeHead(200, {
+        "Content-Type": "application/json",
+        "Access-Control-Allow-Origin": "*",
+      });
+      res.end(JSON.stringify(state));
+    } else if (pathname === "/api/history" && req.method === "GET") {
+      serveHistory(req, res);
+    } else if (pathname.startsWith("/api/history/") && req.method === "GET") {
+      const batchId = decodeURIComponent(pathname.slice("/api/history/".length));
+      serveHistoryEntry(req, res, batchId);
+    } else if (pathname.startsWith("/api/status-md/") && req.method === "GET") {
+      const taskId = decodeURIComponent(pathname.slice("/api/status-md/".length));
+      serveStatusMd(req, res, taskId);
+    } else if (pathname === "/api/preferences" && req.method === "GET") {
+      handleGetPreferences(req, res);
+    } else if (pathname === "/api/preferences" && req.method === "POST") {
+      handlePostPreferences(req, res);
+    } else if (req.method === "OPTIONS") {
+      // CORS preflight for POST
+      res.writeHead(204, {
+        "Access-Control-Allow-Origin": "*",
+        "Access-Control-Allow-Methods": "GET, POST, OPTIONS",
+        "Access-Control-Allow-Headers": "Content-Type",
+      });
+      res.end();
+    } else {
+      serveStatic(req, res);
+    }
+  });
+
+  return server;
+}
+
+// ─── Browser Auto-Open ─────────────────────────────────────────────────────
+
+function openBrowser(url) {
+  const cmd = process.platform === "win32" ? "start"
+    : process.platform === "darwin" ? "open" : "xdg-open";
+  exec(`${cmd} ${url}`, () => {}); // fire-and-forget
+}
+
+// ─── Main ───────────────────────────────────────────────────────────────────
+
+/** Try to listen on a port. Resolves with the port on success, rejects on EADDRINUSE. */
+function tryListen(server, port) {
+  return new Promise((resolve, reject) => {
+    const onError = (err) => {
+      server.removeListener("listening", onListening);
+      reject(err);
+    };
+    const onListening = () => {
+      server.removeListener("error", onError);
+      resolve(port);
+    };
+    server.once("error", onError);
+    server.once("listening", onListening);
+    server.listen(port);
+  });
+}
+
+/** Find an available port starting from `start`, trying up to MAX_PORT_ATTEMPTS. */
+async function findPort(server, start, explicit) {
+  // If the user explicitly passed --port, only try that one
+  if (explicit) {
+    try {
+      return await tryListen(server, start);
+    } catch (err) {
+      if (err.code === "EADDRINUSE") {
+        console.error(`\n  Port ${start} is already in use.`);
+        console.error(`  Try: taskplane dashboard --port ${start + 1}\n`);
+        process.exit(1);
+      }
+      throw err;
+    }
+  }
+  // Auto-scan for an available port
+  for (let port = start; port < start + MAX_PORT_ATTEMPTS; port++) {
+    try {
+      return await tryListen(server, port);
+    } catch (err) {
+      if (err.code === "EADDRINUSE") {
+        // Close the server so we can retry on the next port
+        server.close();
+        server = createServer();
+        continue;
+      }
+      throw err;
+    }
+  }
+  console.error(`\n  No available port found in range ${start}-${start + MAX_PORT_ATTEMPTS - 1}.\n`);
+  process.exit(1);
+}
+
+async function main() {
+  const opts = parseArgs();
+
+  // Resolve project root: --root flag > cwd.
+  // In workspace mode this is the workspace root. All state/sidecar files
+  // (batch-state, lane-state, conversation logs, batch-history) live at
+  // <REPO_ROOT>/.pi/ and are NOT affected by taskplane-pointer.json.
+  // The pointer only redirects config/agent resolution in task-runner and
+  // orchestrator — the dashboard reads only runtime state, so no pointer
+  // resolution is performed here.
+  REPO_ROOT = path.resolve(opts.root || process.cwd());
+  BATCH_STATE_PATH = path.join(REPO_ROOT, ".pi", "batch-state.json");
+  BATCH_HISTORY_PATH = path.join(REPO_ROOT, ".pi", "batch-history.json");
+
+  const server = createServer();
+  const explicitPort = process.argv.slice(2).includes("--port");
+  const port = await findPort(server, opts.port, explicitPort);
+
+  console.log(`\n  Orchestrator Dashboard → http://localhost:${port}\n`);
+
+  // Broadcast state to all SSE clients on interval
+  const pollTimer = setInterval(broadcastState, POLL_INTERVAL);
+
+  // Also watch batch-state.json for immediate push on change
+  try {
+    const batchDir = path.dirname(BATCH_STATE_PATH);
+    if (fs.existsSync(batchDir)) {
+      let debounce = null;
+      fs.watch(batchDir, (eventType, filename) => {
+        if (filename === "batch-state.json") {
+          clearTimeout(debounce);
+          debounce = setTimeout(broadcastState, 200);
+        }
+      });
+    }
+  } catch {
+    // fs.watch not supported — polling is sufficient
+  }
+
+  // Auto-open browser
+  if (opts.open) {
+    setTimeout(() => openBrowser(`http://localhost:${port}`), 500);
+  }
+
+  // Graceful shutdown
+  function cleanup() {
+    clearInterval(pollTimer);
+    for (const client of sseClients) {
+      try { client.end(); } catch {}
+    }
+    server.close();
+    process.exit(0);
+  }
+
+  process.on("SIGINT", cleanup);
+  process.on("SIGTERM", cleanup);
+}
+
+main();