@gotcos/glasses-server 6.30.0 → 6.32.0

package/CHANGELOG.md

@@ -2219,6 +2219,88 @@ unsaved capture, and makes batch status stop lying about finished work.
 
 # Changelog
 
+## [6.32.0] - 2026-08-16
+
+### Continue now blocks on WORKING, not on merely open
+
+Continuing a Claude Code thread from COS used to be refused whenever any foreign
+process held it. That came from `~/.claude/sessions/<pid>.json`, which records an
+OPEN WINDOW, not an agent generating: a session that finished ten minutes ago
+with the window still up looked identical to one mid-turn. Anyone who leaves
+Claude Code windows open was therefore refused on exactly the threads they work
+in, and allowed only on the ones they had abandoned.
+
+- A foreign holder is now terminal only while it is **demonstrably writing**. A
+  holder measured idle (registry record alive, transcript stale past the same
+  30s window `running_active` uses) is continuable.
+- New refusal reason **`native_thread_working`**, distinct from
+  `live_desktop_process`, because the two ask for different things. Working
+  clears itself in seconds and the copy says to wait; `live_desktop_process` now
+  means COS could not measure the thread at all and says so rather than implying
+  the thread is busy.
+- Off by default behind **`COS_THREAD_ATTACH_IDLE_HOLDER=1`**. Unset, no
+  transcript clock is wired, every foreign holder reads unknown, and the gate
+  behaves exactly as it did in 6.31.0. Composes with `COS_THREAD_ATTACH_ENABLED`.
+
+Verified against a real interactive `claude` 2.1.229 held open and idle before
+the gate was touched. Writing into it delivers normally and never corrupts:
+across six injections and six desktop turns, including three writers landing
+within 45ms of each other, every transcript prefix stayed byte-identical, with
+no unparseable lines and no dangling parent references. What it DOES do is fork
+the conversation: the holder's in-memory view is stale, so its next turn branches
+off the pre-injection tail and the two writers stop seeing each other. Nothing is
+lost, and the divergence is caught one layer up by the head watermark, which
+changes on a desktop write and refuses the next COS turn with
+`native_thread_changed`. The full canary, including what it does NOT license, is
+recorded in `server/lib/thread-occupancy.ts` under THE IDLE-HOLDER RELAXATION.
+
+## [6.31.0] - 2026-08-16
+
+### "Running" meant an open window, not a working agent
+
+- `GET /api/agent-sessions` adds **`running_active`**: the thread's transcript was
+  written in the last 30 seconds, so an agent is generating in it right now.
+- This exists because `running` could not answer that question and was being read
+  as though it could. It comes from a process registry, and a Claude Code record
+  describes an open WINDOW (`kind: interactive`, `entrypoint: claude-desktop`) —
+  so a session that finished an hour ago, with the window still up, stayed
+  `running: true` forever. A finished session reported itself active on the
+  glasses and never cleared.
+- The transcript is the only part of a session with a clock in it. While a
+  session generates, its jsonl mtime tracks the wall clock to within a second;
+  when it stops, the mtime goes stale while the registry record does not. That
+  divergence is the whole signal.
+- The window is 30 seconds, deliberately much wider than the observed write
+  cadence, so a long tool call between writes does not flap a working session to
+  idle and back.
+- **Costs one stat per HELD thread, not per listed thread.** Freshness is layered
+  on after the occupancy scan and only over what it found, so a list with nothing
+  running does no extra filesystem work at all.
+- Anything unmeasurable — no transcript, unreadable file, a timestamp from the
+  future — reads as NOT active. The row then falls back to "open", which is still
+  true of a thread with a live owner. The stronger claim has to be earned by an
+  actual observation.
+- **Still a display hint and still never a write gate.** `attachability` and
+  `attach` keep probing at the moment of the write, unchanged.
+
+### The session detail describes its own liveness
+
+- `GET /api/agent-sessions/:provider/:sessionId` now carries the same three
+  `running_*` fields plus `running_stamped: true` and `runningDegraded`.
+- Until now only the LIST stamped occupancy, so the detail page had to borrow the
+  flags from whichever row the user tapped. Those flags froze at the moment of the
+  tap, which is the other half of why a finished session never cleared: the page
+  had no way to ask again. A polling client needs a payload that carries its own
+  liveness.
+- `running_stamped` exists so a client can tell an old server's silence from a new
+  server's `false`. They demand opposite behaviour: keep the borrowed hint, or drop
+  it as stale.
+- Nearly free: the handler already resolves and stats the transcript, so only the
+  single-thread occupancy scan is new, and that is the same scan `attachability`
+  runs on every menu open.
+
+**Required by the COS Glasses build that ships the honest activity line.**
+
 ## [6.30.0] - 2026-08-16
 
 ### Sessions know which threads are live

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@gotcos/glasses-server",
-  "version": "6.30.0",
+  "version": "6.32.0",
   "description": "COS Glasses \u2014 self-hosted AI heads-up-display server for Even G2 smart glasses, powered by Claude Code, Codex, or Cursor Agent CLI",
   "type": "module",
   "bin": {

package/server/index.ts

@@ -23,7 +23,12 @@ import { claudeSessionsRouter } from './routes/claude-sessions.js'
 import { createAgentSessionBindingsRouter } from './routes/agent-session-bindings.js'
 import { AgentSessionBindingRegistry } from './lib/agent-session-binding-registry.js'
 import { cosSpawnedPids } from './lib/agent-session-ownership-store.js'
-import { realOccupancyDirs, realOccupancyProbes } from './lib/occupancy-probes.js'
+import {
+  idleHolderContinueEnabled,
+  realOccupancyDirs,
+  realOccupancyProbes,
+  withTranscriptClock,
+} from './lib/occupancy-probes.js'
 import { realAttachedWorkspaceDeps, resolveAttachedWorkspace } from './lib/attached-workspace.js'
 import { deliverAttachedTurn, realAttachedTurnDeps } from './lib/attached-provider-adapter.js'
 import { forkThread, realForkDeps } from './lib/fork-thread.js'
@@ -301,10 +306,28 @@ bindingReapTimer.unref()
 
 // Built once. Each of these reads the disk, so sharing them keeps an attach from
 // re-deriving roots per request.
-const occupancyProbes = realOccupancyProbes(cosSpawnedPids)
 const occupancyDirs = realOccupancyDirs()
 const nativeHeadDeps = realNativeHeadDeps()
 const attachedWorkspaceDeps = realAttachedWorkspaceDeps(nativeHeadDeps)
+/**
+ * THE ONE PLACE the idle-holder relaxation is switched on (6.32.0).
+ *
+ * Without the clock, a foreign holder is terminal and Continue is refused for
+ * any thread with a Claude Code window open on it — including the ones Miles
+ * actually works in, because he leaves those windows open. With it, a holder
+ * measured idle is continuable and only a holder measured WRITING is refused.
+ *
+ * Both gates read this same object — the route's detector via `probes` below,
+ * and the adapter's pre-spawn preflight via `occupancyProbes` at the closure
+ * further down — so they cannot come to different conclusions. That is the whole
+ * reason the clock is a probe rather than a check bolted onto each call site.
+ *
+ * Read the canary evidence in `thread-occupancy.ts` under THE IDLE-HOLDER
+ * RELAXATION before changing this line.
+ */
+const occupancyProbes = idleHolderContinueEnabled()
+  ? withTranscriptClock(realOccupancyProbes(cosSpawnedPids), nativeHeadDeps)
+  : realOccupancyProbes(cosSpawnedPids)
 
 /**
  * The shim between the route's request shape and the adapter's.

package/server/lib/occupancy-probes.ts

@@ -49,6 +49,7 @@ import {
 import { homedir } from 'node:os'
 import { basename, join, resolve } from 'node:path'
 import { NATIVE_THREAD_ID_RE } from './native-thread-id.js'
+import { transcriptPathFor, type NativeHeadDeps } from './native-head.js'
 import { parseProcStartUtcMs, type OccupancyDirs, type OccupancyProbes } from './thread-occupancy.js'
 
 // Re-exported rather than reimplemented. `claudeSessionsDir` already encodes the
@@ -440,5 +441,61 @@ export function realOccupancyProbes(ledger: SpawnLedgerAccessor): OccupancyProbe
     readFile,
     lockHolders,
     cosSpawnedPids: () => sanitizeLedger(ledger()),
+    // transcriptMtimeMs is DELIBERATELY absent here. See `withTranscriptClock`.
+  }
+}
+
+/**
+ * Is the idle-holder relaxation switched on for this install?
+ *
+ * ONE READER, exported so every other module imports it rather than re-reading
+ * `process.env` — the same anti-drift rule `threadAttachEnabled` follows, and for
+ * the same reason: a second copy is how a surface comes to advertise a write path
+ * the gate will refuse.
+ *
+ * `=== '1'` opt-IN, not `!== '0'` opt-out. This decides whether COS may write
+ * into a conversation a human still has open, so anything ambiguous — unset,
+ * empty, 'true', 'yes' — reads as OFF. It composes with
+ * `COS_THREAD_ATTACH_ENABLED`: with attach itself off, this flag does nothing at
+ * all, because there is no write path to relax.
+ */
+export function idleHolderContinueEnabled(): boolean {
+  return process.env.COS_THREAD_ATTACH_IDLE_HOLDER === '1'
+}
+
+/**
+ * Add a transcript clock to a probe set, so an idle foreign holder can be told
+ * apart from a working one.
+ *
+ * WHY THIS IS A SEPARATE WRAPPER rather than a member of `realOccupancyProbes`:
+ * the strict gate has to be what you get by DEFAULT. A probe set built the
+ * ordinary way has no clock, every foreign holder reads `unknown`, and
+ * `threadOccupancy` refuses exactly as it did before 6.32.0. Turning the
+ * relaxation on is then one visible call at one wiring site, and turning it off
+ * is deleting that call — no rollback, no second code path to keep in step.
+ *
+ * Null on every failure: an unresolvable id, a missing file, a stat that threw.
+ * `holderActivity` reads null as `unknown`, which refuses. There is deliberately
+ * no path from "could not read the transcript" to "the holder is idle".
+ *
+ * `statSync` and not `stat`: `OccupancyProbes` is synchronous throughout, and the
+ * adapter's pre-spawn preflight REFUSES a thenable rather than awaiting one —
+ * awaiting there would reopen the very race the check exists to close.
+ */
+export function withTranscriptClock(
+  probes: OccupancyProbes,
+  headDeps: NativeHeadDeps,
+): OccupancyProbes {
+  return {
+    ...probes,
+    transcriptMtimeMs: (provider, threadId) => {
+      try {
+        const path = transcriptPathFor(provider, threadId, headDeps)
+        if (path === null) return null
+        return statSync(path).mtimeMs
+      } catch {
+        return null
+      }
+    },
   }
 }

package/server/lib/occupied-threads.ts

@@ -32,6 +32,7 @@ import {
   claudeOwners,
   codexLockPath,
   codexOwners,
+  isActiveRecently,
   type OccupancyDirs,
   type OccupancyProbes,
   type ThreadOwner,
@@ -41,10 +42,15 @@ import { isValidNativeThreadId } from './native-thread-id.js'
 export interface OccupiedThread {
   threadId: string
   /**
-   * Every process working in this thread, COS's own children included.
+   * Every process holding this thread open, COS's own children included.
    *
-   * This is the DISPLAY question — "is an agent working in here right now" — and
-   * a turn COS queued is just as much a running agent as a desktop window is.
+   * WHAT THIS ACTUALLY MEASURES, stated plainly because the first version of the
+   * feature got it wrong in user-facing copy: a registry record means A PROCESS
+   * HOLDS THIS THREAD OPEN. It does NOT mean an agent is generating. A Claude
+   * Code record reads `kind: interactive, entrypoint: claude-desktop` for an open
+   * WINDOW, so a session that finished ten minutes ago with the window still up
+   * has an owner forever. The lens said "ACTIVE NOW" off this number and was
+   * lying; it now says the thread is OPEN, which is all this count can support.
    */
   owners: number
   /**
@@ -55,6 +61,16 @@ export interface OccupiedThread {
    * window does.
    */
   foreignOwners: number
+  /**
+   * The transcript was WRITTEN inside the freshness window: an agent is
+   * generating here, not merely holding the file open.
+   *
+   * This is the signal `owners` cannot give. Set only from a positive
+   * observation of a recent write — see `withActiveRecently`, which is where it
+   * is filled in. `occupiedThreads` itself has no path to a transcript and
+   * always leaves it false.
+   */
+  activeRecently: boolean
 }
 
 export interface OccupiedScan {
@@ -178,7 +194,11 @@ export function occupiedThreads(
     // hide COS's own queued turn from the very screen the user opens to watch it.
     const foreign = owners.filter(o => !o.selfOwned).length
     if (owners.length > 0) {
-      occupied.set(threadId, { threadId, owners: owners.length, foreignOwners: foreign })
+      // `activeRecently` is FALSE here and can only be raised by
+      // `withActiveRecently`. This function reads a process registry; a registry
+      // record cannot tell generating from idling, and inferring one from the
+      // other is the exact defect this field exists to fix.
+      occupied.set(threadId, { threadId, owners: owners.length, foreignOwners: foreign, activeRecently: false })
     }
   }
 
@@ -189,3 +209,48 @@ export function occupiedThreads(
 export function noOccupancyKnown(): OccupiedScan {
   return { occupied: new Map(EMPTY.occupied), degraded: true }
 }
+
+/**
+ * The freshness window and the raw reading now live in `thread-occupancy.ts`.
+ *
+ * MOVED, not duplicated (6.32.0). The write gate needs the same window this
+ * hint uses, and it is the lower module, so keeping a second copy here would be
+ * two definitions of "recently" that could drift. Re-exported so every existing
+ * importer of these two names is unaffected.
+ *
+ * Note the polarity difference before reusing `isActiveRecently` anywhere new:
+ * for a HINT, its `false` safely means "render OPEN". For a GATE, that same
+ * `false` would mean "allow the write", and unmeasurable must not mean allowed.
+ * `holderActivity` in `thread-occupancy.ts` is the gate-side reading.
+ */
+export { ACTIVE_RECENTLY_WINDOW_MS, isActiveRecently } from './thread-occupancy.js'
+
+/**
+ * Fill in `activeRecently` for a scan, from transcript write times.
+ *
+ * PURE, and separate from `occupiedThreads` on purpose. Resolving a thread id to
+ * a transcript path is the session STORE's job (it owns the projects/rollouts
+ * layout), and this module owns process occupancy; wiring the store in here
+ * would make an occupancy scan depend on a filesystem it does not otherwise know
+ * about. The caller stats — for HELD THREADS ONLY, which is what keeps this
+ * cheap, since it is one stat per already-occupied row and never one per listed
+ * row — and hands the readings in.
+ *
+ * A thread with no entry in the map is not an error and not "recent": it reads
+ * exactly like an unreadable one, because from here the two are the same
+ * absence of evidence.
+ */
+export function withActiveRecently(
+  scan: OccupiedScan,
+  transcriptMtimes: ReadonlyMap<string, number | null>,
+  nowMs: number,
+): OccupiedScan {
+  const occupied = new Map<string, OccupiedThread>()
+  for (const [threadId, record] of scan.occupied) {
+    occupied.set(threadId, {
+      ...record,
+      activeRecently: isActiveRecently(transcriptMtimes.get(threadId), nowMs),
+    })
+  }
+  return { occupied, degraded: scan.degraded }
+}

package/server/lib/thread-occupancy.ts

@@ -56,6 +56,12 @@ export interface ThreadOwner {
 
 export type OccupancyReason =
   | 'live_desktop_process'
+  // A foreign holder that is DEMONSTRABLY generating right now, as opposed to
+  // merely holding the thread open. Separate from `live_desktop_process` because
+  // the two ask the user for different things: this one clears by itself in
+  // seconds and is worth waiting out, the other needs a window closed. Only
+  // reachable when a transcript clock is wired — see THE IDLE-HOLDER RELAXATION below.
+  | 'native_thread_working'
   | 'unsupported_provider'
   | 'invalid_thread_id'
   | 'detector_unavailable'
@@ -73,13 +79,30 @@ export interface Occupancy {
   owners: ThreadOwner[]
   /** Null only when attachable. Drives the Control/lens footer copy. */
   reason: OccupancyReason | null
+  /**
+   * This verdict is attachable DESPITE a foreign owner, because that owner was
+   * measured idle. See THE IDLE-HOLDER RELAXATION below.
+   *
+   * It exists so the permissive outcome has to be DECLARED rather than inferred
+   * from `attachable` alone. `projectAttachability` treats a foreign owner on an
+   * attachable verdict as a contradiction and forces it back to refused; without
+   * a positive marker, relaxing the gate would mean deleting that check, and the
+   * check is what catches a future detector that flips `attachable` by accident.
+   * With the marker, the only way to reach the permissive path is to say so.
+   *
+   * Absent (not false) on every other verdict, so `=== true` is the only test.
+   */
+  idleHolder?: true
 }
 
 /**
  * What a scan could not establish. Any non-null value forbids attaching, even
  * with zero owners — that is the whole point of the type.
  */
-type Doubt = Exclude<OccupancyReason, 'live_desktop_process' | 'unsupported_provider' | 'invalid_thread_id'> | null
+type Doubt = Exclude<
+  OccupancyReason,
+  'live_desktop_process' | 'native_thread_working' | 'unsupported_provider' | 'invalid_thread_id'
+> | null
 
 interface ScanResult {
   owners: ThreadOwner[]
@@ -111,6 +134,127 @@ export interface OccupancyProbes {
    * membership — the start time is what makes the claim checkable.
    */
   cosSpawnedPids: () => ReadonlyMap<number, number>
+  /**
+   * Transcript write time for this thread, epoch ms, or null when it cannot be
+   * read. OPTIONAL, and its ABSENCE is the default.
+   *
+   * Absent means this install has no clock on the thread, which means every
+   * foreign holder stays `live_desktop_process` — byte-for-byte the behaviour
+   * that shipped before the idle-holder relaxation existed. That is why it is
+   * optional rather than required: the strict gate is what you get by doing
+   * nothing, and the relaxation has to be wired on purpose (see `server/index.ts`,
+   * where the wiring is itself behind `COS_THREAD_ATTACH_IDLE_HOLDER`).
+   *
+   * Null is NOT idle. See `holderActivity` for why that distinction is the whole
+   * safety property here.
+   */
+  transcriptMtimeMs?: (provider: OccupancyProvider, threadId: string) => number | null
+}
+
+/**
+ * How recently the transcript must have been written for a held thread to read
+ * as WORKING rather than merely OPEN.
+ *
+ * WHY A SECOND SIGNAL EXISTS AT ALL. `owners` answers "a process holds this
+ * thread", which is not the question the user is asking. He watched a session
+ * finish on his Mac while the glasses still showed it active, because the window
+ * was still open and the registry record therefore still existed. Occupancy has
+ * no clock in it; the transcript does.
+ *
+ * MEASURED, not assumed: while a session generates, its jsonl mtime tracks the
+ * wall clock to within a second, and when generation stops the mtime goes stale
+ * while the registry record stays exactly where it was. That divergence is the
+ * whole signal.
+ *
+ * 30 seconds is deliberately far wider than the observed sub-second write
+ * cadence. The gap this has to survive is a long tool call or a slow first token
+ * between writes, and a window sized to the cadence would flap a working session
+ * to OPEN and back every time the model paused to think. The cost of the wide
+ * window is bounded and known: a session that stops is reported working for up
+ * to 30 more seconds, which is a late correction rather than a permanent lie.
+ *
+ * LIVES HERE, not in `occupied-threads.ts`, since 6.32.0. The display hint and
+ * the write gate must agree on what "recently" means, and the gate is the lower
+ * module of the two. `occupied-threads.ts` re-exports it, so every existing
+ * importer is unaffected.
+ */
+export const ACTIVE_RECENTLY_WINDOW_MS = 30_000
+
+/**
+ * Was this transcript written inside the window?
+ *
+ * NULL IS NOT "RECENT". An unresolvable path, an unreadable file, a stat that
+ * threw — every one of them arrives here as null and answers false, so the row
+ * falls back to OPEN. That is not a fail-open FOR A DISPLAY HINT: OPEN is still
+ * a true statement about a thread with a live owner. It is the STRONGER claim,
+ * "an agent is working in here", that has to be earned by an actual observation.
+ *
+ * A timestamp far in the FUTURE is refused for the same reason rather than
+ * treated as maximally fresh. Skew of a few seconds is normal and lands inside
+ * the window; a file dated next week is a clock this server cannot reason about,
+ * and letting it manufacture a permanent "working" badge would rebuild the bug
+ * from the other direction.
+ *
+ * DO NOT CALL THIS FROM A WRITE GATE. Its false has two meanings — "measured,
+ * and stale" and "could not measure" — and a gate that treats the second as the
+ * first is fail-open. `holderActivity` is the gate-side reading.
+ */
+export function isActiveRecently(mtimeMs: number | null | undefined, nowMs: number): boolean {
+  if (typeof mtimeMs !== 'number' || !Number.isFinite(mtimeMs)) return false
+  if (!Number.isFinite(nowMs)) return false
+  return Math.abs(nowMs - mtimeMs) <= ACTIVE_RECENTLY_WINDOW_MS
+}
+
+/**
+ * What a foreign holder is doing, for a caller that must DECIDE rather than
+ * render.
+ *
+ * THREE VALUES, AND THE THIRD IS THE POINT. `isActiveRecently` collapses "I
+ * measured a stale file" and "I could not measure anything" into the same
+ * `false`, which is correct for a badge — the weaker claim, OPEN, is true either
+ * way — and catastrophic for a gate, where that same `false` would mean ALLOW
+ * THE WRITE. Same reading, opposite polarity, so the gate gets its own function
+ * instead of reusing the hint's boolean.
+ *
+ * Only `idle` is a positive observation of an idle holder: a real number, not in
+ * the future, measured outside the window. Everything else is `unknown` and
+ * refuses. This is the same "no owner found is not no owner" rule the rest of
+ * this module runs on, applied to the clock instead of the registry.
+ *
+ * The window itself is NOT redefined here — `isActiveRecently` decides what
+ * recent means, so the badge and the gate can never drift apart on it.
+ */
+export type HolderActivity = 'working' | 'idle' | 'unknown'
+
+export function holderActivity(mtimeMs: number | null | undefined, nowMs: number): HolderActivity {
+  if (typeof mtimeMs !== 'number' || !Number.isFinite(mtimeMs)) return 'unknown'
+  if (!Number.isFinite(nowMs)) return 'unknown'
+  // A future-dated transcript is a clock this server cannot reason about. The
+  // hint may treat a few seconds of skew as freshness; the gate may not treat
+  // ANY amount of it as idleness, because "stale" is the permissive answer here
+  // and a wrong clock would hand it out forever.
+  if (mtimeMs > nowMs + ACTIVE_RECENTLY_WINDOW_MS) return 'unknown'
+  return isActiveRecently(mtimeMs, nowMs) ? 'working' : 'idle'
+}
+
+/**
+ * Read the transcript clock for a thread, refusing to guess.
+ *
+ * A probe that is absent, throws, or answers null all land on `unknown`, which
+ * refuses. There is no path from a failed reading to a permissive verdict.
+ */
+function readHolderActivity(
+  provider: OccupancyProvider,
+  threadId: string,
+  probes: OccupancyProbes,
+  nowMs: number,
+): HolderActivity {
+  if (typeof probes.transcriptMtimeMs !== 'function') return 'unknown'
+  try {
+    return holderActivity(probes.transcriptMtimeMs(provider, threadId), nowMs)
+  } catch {
+    return 'unknown'
+  }
 }
 
 /** A Claude registry filename is exactly `<pid>.json`. Not `*.json`. */
@@ -321,14 +465,81 @@ export interface OccupancyDirs {
   codexLocksDir: string
 }
 
+// ===========================================================================
+// THE IDLE-HOLDER RELAXATION (6.32.0)
+// ===========================================================================
+//
+// This loosens a guard that was deliberate. Read this before touching it, and
+// do not widen it further without repeating the experiment.
+//
+// WHAT CHANGED. A foreign owner used to be terminal on its own. It is now
+// terminal only while that owner is DEMONSTRABLY WRITING. A holder measured
+// idle — registry record alive, transcript stale — is continuable.
+//
+// WHY THE OLD RULE WAS WRONG. The registry records an OPEN WINDOW, not active
+// generation. `~/.claude/sessions/<pid>.json` for a session that finished ten
+// minutes ago is byte-identical to one mid-turn. Miles keeps Claude Code
+// windows open, so Continue was refused for exactly the threads he cares about
+// and allowed only for the ones he had abandoned. The gate was measuring the
+// wrong thing, not measuring it too strictly.
+//
+// THE CANARY, run 2026-08-16 against a real interactive `claude` 2.1.229 held
+// open in tmux (pid 48446, `kind: interactive`, `entrypoint: claude-desktop`,
+// indistinguishable from a desktop window) in a scratch workspace. Turns were
+// injected with the EXACT argv this server spawns. Raw findings:
+//
+//   1. DELIVERY WORKS. `claude -p --resume <id>` into an idle-held thread
+//      exits 0 in ~5s and returns its result. Same session id, same file.
+//   2. NO CORRUPTION. Across 6 injections and 6 desktop turns, including
+//      three SDK writers landing within 45ms of each other on the same node,
+//      the transcript stayed append-only: every prefix byte-identical before
+//      and after (`cmp`), zero unparseable lines, zero dangling parentUuid.
+//      Claude Code appends whole rows, so byte interleaving does not occur.
+//   3. THE HOLDER IGNORES THE INJECTED TURN — this is the real finding. Its
+//      in-memory view is stale and stays stale. Asked afterwards to list every
+//      word it had been told to reply with, it answered "ALPHA": the injected
+//      "BRAVO" was invisible to it.
+//   4. NOTHING IS CLOBBERED, BUT THE THREAD FORKS. The holder's next turn
+//      parented to the PRE-INJECTION tail, making the transcript a tree. Both
+//      turns survive in full; they are on different branches, and from then on
+//      each writer sees only its own. A later `--resume` read back
+//      "ALPHA, BRAVO" and never saw the desktop's "CHARLIE".
+//   5. THE WORKING CASE IS DIFFERENT AND STAYS BLOCKED. With the holder
+//      generating, the transcript mtime was 6.2s old and `holderActivity`
+//      returned `working`, so this relaxation does not fire there at all.
+//
+// SO THE ANSWER IS: no clobbering, no corruption, no data loss — but the two
+// views diverge silently after the write, and neither side is told by THIS
+// module. What makes that acceptable is the guard one layer up: `nativeHead`
+// digests the transcript tail, and a desktop write CHANGES it (measured:
+// nh1:faba1b35... -> nh1:ace93a1f...), so the next COS turn on that binding is
+// refused with `native_thread_changed` and the user is offered refresh, continue
+// anyway, or fork. Divergence is caught by the CONTENT watermark, which is the
+// signal that can actually see it. Occupancy never could.
+//
+// WHAT THIS IS NOT A LICENCE FOR. Do not extend the same reasoning to the
+// `doubt` reasons below: those mean the scan could not SEE, and an unreadable
+// registry is not an idle holder. Do not relax the `working` branch on the
+// grounds that "nothing got corrupted in the canary either" — case 5 was never
+// run to completion precisely because it stays blocked. And do not reach for
+// `isActiveRecently` here; its `false` means "stale OR unmeasurable", and only
+// `holderActivity` separates those.
+//
+// TURNING IT OFF. Unwire `probes.transcriptMtimeMs` (in practice: unset
+// `COS_THREAD_ATTACH_IDLE_HOLDER`). Every foreign holder then reads `unknown`
+// and refuses, which is the pre-6.32.0 gate exactly. No rollback needed.
+
 /**
  * The Phase 0 attach precondition.
  *
  * Returns attachable ONLY when a supported provider proved its detector exists,
- * read every candidate record, and found no foreign owner. Every other outcome
- * names why. The whole function is wrapped: a throwing probe — including the
- * spawn ledger, which is the most safety-critical of them — is `probe_failed`,
+ * read every candidate record, and found no owner it must respect. Every other
+ * outcome names why. The whole function is wrapped: a throwing probe — including
+ * the spawn ledger, which is the most safety-critical of them — is `probe_failed`,
  * never an exception escaping into a route.
+ *
+ * A foreign owner measured IDLE is the one exception, and it is marked
+ * `idleHolder` on the way out. See THE IDLE-HOLDER RELAXATION above.
  */
 export function threadOccupancy(
   provider: string,
@@ -358,10 +569,23 @@ export function threadOccupancy(
 
   const foreign = result.owners.filter(o => !o.selfOwned)
   if (foreign.length > 0) {
-    return { attachable: false, owners: result.owners, reason: 'live_desktop_process' }
+    const activity = readHolderActivity(provider, threadId, probes, Date.now())
+    if (activity !== 'idle') {
+      return {
+        attachable: false,
+        owners: result.owners,
+        reason: activity === 'working' ? 'native_thread_working' : 'live_desktop_process',
+      }
+    }
+    // FALL THROUGH, deliberately. See THE IDLE-HOLDER RELAXATION below for why this is
+    // safe and what it is NOT. Everything after this point still applies: a scan
+    // that could not establish something still refuses on the next line.
   }
   if (result.doubt !== null) {
     return { attachable: false, owners: result.owners, reason: result.doubt }
   }
+  if (foreign.length > 0) {
+    return { attachable: true, owners: result.owners, reason: null, idleHolder: true }
+  }
   return { attachable: true, owners: result.owners, reason: null }
 }

package/server/routes/agent-session-bindings.ts

@@ -391,8 +391,17 @@ export const REASON_COPY: Record<OccupancyReason, string> = {
   // supported permanent configuration, not a degraded one (plan 4.9).
   attach_disabled:
     'Continuing a thread on your Mac is turned off. COS is read-only here. Fork it instead.',
+  // Held open by another app AND COS has no clock on the thread, so it cannot
+  // tell working from idle. Since 6.32.0 an idle holder is continuable, which
+  // makes this the "could not measure" case rather than the "someone else is
+  // here" case, and the copy says which.
   live_desktop_process:
-    'Open on your Mac. COS will not write into a thread another app is holding. Fork it instead.',
+    'Open on your Mac, and COS cannot tell whether it is still working. It will not write into it. Fork it instead.',
+  // Measured, and the answer was yes. Deliberately a different instruction from
+  // the line above: this clears on its own within seconds, so the useful advice
+  // is to wait, with fork as the fallback rather than the recommendation.
+  native_thread_working:
+    'Your Mac is writing to this thread right now. Wait a few seconds and try again, or fork it.',
   unsupported_provider:
     'This assistant cannot be continued from COS yet. Fork it instead.',
   invalid_thread_id:
@@ -672,14 +681,26 @@ export interface AttachabilityBody {
  * permissive way if simply forwarded: attachable with a reason, attachable with an
  * owner that is not provably ours, and a non-array owners field. Any of them is a
  * defect upstream, and a defect must not resolve to permissive.
+ *
+ * THE ONE EXEMPTION, added with the idle-holder relaxation in 6.32.0: a foreign
+ * owner is allowed on an attachable verdict when the verdict itself carries
+ * `idleHolder === true`. The check is not weakened by this, it is made explicit —
+ * before, "attachable" and "no foreign owner" were the same claim, so a detector
+ * that flipped `attachable` by mistake was caught here. It still is. What can no
+ * longer be caught here is a detector that ALSO sets `idleHolder`, which takes a
+ * deliberate edit in `thread-occupancy.ts` rather than an accident, and which the
+ * mutation tests over that file cover.
  */
 export function projectAttachability(verdict: Occupancy): AttachabilityBody {
   const owners = Array.isArray(verdict?.owners) ? verdict.owners : null
+  // Strictly `=== true`. An idleHolder of 1, 'yes', or {} is a malformed verdict,
+  // and a malformed verdict must not buy an exemption.
+  const foreignOwnerDeclared = verdict?.idleHolder === true
   const sound =
     verdict?.attachable === true &&
     verdict.reason === null &&
     owners !== null &&
-    owners.every(owner => owner?.selfOwned === true)
+    owners.every(owner => owner?.selfOwned === true || foreignOwnerDeclared)
   const reason: OccupancyReason | null = sound ? null : ((verdict?.reason ?? 'probe_failed') as OccupancyReason)
   return {
     attachable: sound,

package/server/routes/agent-sessions.ts

@@ -32,7 +32,13 @@ import {
 import { searchAgentSessions, type AgentSessionSearchHit } from '../lib/agent-session-search.js'
 import { claudeSessionNamesVisible, claudeSessionsDir, claudeSessionsEnabled, readClaudePeers } from './claude-sessions.js'
 import { workspaceFromCwd } from '../lib/claude-session-registry.js'
-import { occupiedThreads, noOccupancyKnown, type OccupiedScan, type OccupiedThread } from '../lib/occupied-threads.js'
+import {
+  occupiedThreads,
+  noOccupancyKnown,
+  withActiveRecently,
+  type OccupiedScan,
+  type OccupiedThread,
+} from '../lib/occupied-threads.js'
 import { realOccupancyDirs, realOccupancyProbes } from '../lib/occupancy-probes.js'
 import { cosSpawnedPids } from '../lib/agent-session-ownership-store.js'
 
@@ -105,16 +111,115 @@ function runningThreads(rows: readonly AgentSessionRow[]): OccupiedScan {
   }
 }
 
-/** Stamp the running hint onto a projected row. */
-function withRunning<T extends { session_id: string }>(entry: T, scan: OccupiedScan) {
+/**
+ * When each HELD thread's transcript was last written.
+ *
+ * ONLY the threads the scan already found occupied. That is the entire cost
+ * argument: the occupied set is typically one or two rows out of sixty, so this
+ * is one or two path resolutions and stats per request, not sixty. An empty scan
+ * does no filesystem work at all.
+ *
+ * WHY NOT REUSE `row.modified`, WHICH IS ALREADY AN MTIME. Because for the rows
+ * that matter it is not one. A live Claude row comes from `liveClaudeRows`,
+ * whose `modified` is the registry's `lastActiveAt` (or, absent that, literally
+ * `new Date()`), and `enrichLiveClaude` does not replace it with the file's
+ * mtime. Reading freshness off that field would report an open window as a fresh
+ * write on every single request, which is precisely the bug being fixed.
+ *
+ * Every failure lands on null, and null is doubt rather than "recent" — see
+ * `isActiveRecently`. A row that cannot be measured reads OPEN, which is still
+ * true of a thread with a live owner.
+ */
+async function transcriptMtimes(
+  scan: OccupiedScan,
+  rows: readonly AgentSessionRow[],
+): Promise<Map<string, number | null>> {
+  const mtimes = new Map<string, number | null>()
+  if (scan.occupied.size === 0) return mtimes
+
+  const providerById = new Map<string, AgentProvider>()
+  for (const row of rows) providerById.set(row.session_id, row.provider)
+  const roots = agentSessionRoots()
+
+  await Promise.all([...scan.occupied.keys()].map(async threadId => {
+    try {
+      const provider = providerById.get(threadId)
+      // An occupied id with no row cannot happen today (the scan is built FROM
+      // the rows), but guessing a provider to go looking for a file is how a
+      // caller-supplied id reaches the filesystem. Unknown stays unmeasured.
+      if (!provider) {
+        mtimes.set(threadId, null)
+        return
+      }
+      const file = await findAgentSessionFile(provider, threadId, roots)
+      mtimes.set(threadId, file ? (await stat(file)).mtimeMs : null)
+    } catch {
+      mtimes.set(threadId, null)
+    }
+  }))
+  return mtimes
+}
+
+/**
+ * Stamp the running hint onto a projected row.
+ *
+ * EXPORTED so the WIRE KEYS are covered by a behaviour test rather than by
+ * reading this file. The client reads `running_active` by exact name
+ * (`session-running.ts` accepts only a literal `true` on a known key), so a
+ * rename here is silent on this side and renders every session as merely open on
+ * the lens. Pure: it takes the scan, it does no I/O.
+ */
+export function withRunning<T extends { session_id: string }>(entry: T, scan: OccupiedScan) {
   const occ = scan.occupied.get(entry.session_id)
   return {
     ...entry,
-    // An agent is working in this thread right now, whoever started it.
+    // A process HOLDS this thread open, whoever started it. Not "is generating":
+    // a Claude record for an open window outlives the work by however long the
+    // window stays up, which is why `running_active` exists below.
     running: occ !== undefined,
     // Held by something that is not COS, so a Continue would be refused. The
     // badge reads `running`; the Continue affordance reads this.
     running_foreign: (occ?.foreignOwners ?? 0) > 0,
+    // The transcript was WRITTEN in the last few seconds: an agent is generating
+    // here, not just holding the file. This is the only one of the three that
+    // clears on its own when the work stops. Still a display hint, never a gate.
+    running_active: occ?.activeRecently === true,
+  }
+}
+
+/**
+ * Occupancy plus freshness for ONE thread, for the detail route.
+ *
+ * WHY THE DETAIL ROUTE STAMPS THIS AT ALL. It did not, and the lens had to borrow
+ * the hint from whichever list row the user tapped. That worked exactly once: the
+ * detail page was a single fetch, so the borrowed flags were frozen at the moment
+ * it opened and could never clear, which is half of why a finished session stayed
+ * "active" on the glasses until Miles navigated away. A page that polls needs a
+ * payload that carries its own liveness.
+ *
+ * NEARLY FREE HERE. The handler has already resolved the transcript path and
+ * stat'ed it, so freshness costs nothing extra and only the single-thread
+ * occupancy scan is new — the same scan `attachability` already runs on every
+ * menu open. Contrast the list, where the same work would be sixty scans.
+ *
+ * STILL A DISPLAY HINT. Nothing about being on the detail payload makes it a
+ * gate; `attach` re-probes at the write, unchanged.
+ */
+function runningForThread(provider: AgentProvider, threadId: string, mtimeMs: number): OccupiedScan {
+  if (provider !== 'claude' && provider !== 'codex') return { occupied: new Map(), degraded: false }
+  try {
+    const scan = occupiedThreads(
+      provider,
+      [threadId],
+      realOccupancyProbes(cosSpawnedPids),
+      realOccupancyDirs(),
+    )
+    return withActiveRecently(scan, new Map([[threadId, mtimeMs]]), Date.now())
+  } catch (error) {
+    // The transcript is the point; occupancy is decoration. A probe failure must
+    // never cost the user their session.
+    console.error(`[agent-sessions] detail occupancy failed: ${error instanceof Error ? error.message : error}`)
+    return noOccupancyKnown()
   }
 }
 
@@ -164,7 +269,11 @@ agentSessionsRouter.get('/agent-sessions', async (req, res) => {
     const sort = asSort(req.query.sort)
     const live = await liveClaudeRows()
     const sessions = await listAgentSessions(agentSessionRoots(), new Date(), live, limit, sort)
-    const running = runningThreads(sessions)
+    const scan = runningThreads(sessions)
+    // Freshness is layered on AFTER occupancy, and only over what occupancy
+    // found. `Date.now()` is read once so every row in a payload is judged
+    // against the same instant.
+    const running = withActiveRecently(scan, await transcriptMtimes(scan, sessions), Date.now())
     res.json({
       sessions: sessions.map(row => withRunning(toEntry(row), running)),
       total: sessions.length,
@@ -234,8 +343,19 @@ agentSessionsRouter.get('/agent-sessions/:provider/:sessionId', async (req, res)
       if (named) parsed.display_label = named
     }
     const modified = st.mtime.toISOString()
+    const running = runningForThread(provider, parsed.session_id, st.mtimeMs)
     res.json({
-      session_id: parsed.session_id,
+      ...withRunning({ session_id: parsed.session_id }, running),
+      // The client must be able to tell "this server stamped nothing" from "this
+      // server stamped false", because the two demand opposite behaviour: an old
+      // server's silence means keep using the hint borrowed from the list row, and
+      // a new server's `false` means the thread genuinely went quiet and the
+      // borrowed hint is the stale thing. Without this marker the poll could never
+      // clear the flag it exists to clear.
+      running_stamped: true,
+      // Same meaning as on the list: a probe could not see clearly, so render
+      // unknown rather than treating a quiet scan as "nothing is running".
+      runningDegraded: running.degraded,
       provider: parsed.provider,
       slug: parsed.session_id,
       custom_title: parsed.display_label,