@bybrawe/opencode-loop 0.5.7 → 0.5.9

package/CHANGELOG.md

@@ -1,5 +1,22 @@
 # Changelog
 
+## 0.5.9
+
+- Fixed stale-busy recovery for TUI sessions. Cached `busy` / `retry` status is now short-lived, so due jobs no longer get stuck at `due in every idle` after `/loop-status` or other short custom-command turns.
+- Improved due-timer behavior when OpenCode misses or delays the final idle event after a command turn.
+- Fixed `/compact` routing for current OpenCode TUI builds by trying `session_compact` first, then `session.compact`, then the session summarize fallback.
+- Kept the command markdown templates short; OpenCode still creates a tiny `OK` turn for custom commands because markdown commands are prompt templates.
+
+## 0.5.8
+
+- Added a real due timer for TUI loop jobs, so delayed jobs can wake without waiting for another OpenCode event.
+- Fixed `/loop 1m --no-now ...` style jobs so the first delayed run can fire on time when OpenCode is idle.
+- Fixed scheduled action types such as `/loop-ask`, `/loop-command`, and `/loop-shell` so 0s and delayed jobs are scheduled even when their default mode is `--no-now`.
+- Added busy retry behavior: if a due timer expires while OpenCode is busy, the job waits and retries instead of interrupting the active turn.
+- Reschedules due timers after add, stop, pause, resume, run completion, goal pause/resume/clear, and post-run verification.
+- Shortened command markdown templates so unavoidable custom command turns reply `OK` instead of producing misleading status text.
+- Kept Experimental Goal Mode marked experimental.
+
 ## 0.5.7
 
 - Added experimental Goal Mode with `/loop-goal <objective>`.

package/README.md

@@ -11,7 +11,7 @@ NPM package name: **@bybrawe/opencode-loop**
 
 ## Current status
 
-**v0.5.7 adds experimental Goal Mode.** OpenCode Loop still fixes the recent OpenCode update compatibility issue and keeps the clear action types from v0.5.6. The TUI loop waits for the session to become idle, debounces idle events, and avoids starting a new run while OpenCode is still busy or queued. It separates prompt loops, scheduled question loops, OpenCode slash-command loops, shell loops, compact loops, and now experimental persistent goals so commands like `/compact` are not sent as normal chat text.
+**v0.5.9 is a reliability hotfix for stale-busy status and `/compact` routing.** It keeps the real due timer from v0.5.8, but no longer trusts a cached `busy` / `retry` session status forever. This fixes the case where `/loop-status` or another short custom-command turn could leave a loop showing `due in every idle` without firing. It also routes scheduled `/compact` through OpenCode's current TUI alias first, with fallback paths for older builds.
 
 The known update-related symptoms from older builds are fixed:
 
@@ -22,9 +22,60 @@ The known update-related symptoms from older builds are fixed:
 - `/compact` accidentally being treated as a normal agent prompt
 - intermittent `Tool execution aborted` behavior caused by triggering a new turn too early
 - experimental `/loop-goal` support for goal-driven work that continues until complete, blocked, paused, or cleared
+- `--no-now` interval jobs not waking until another OpenCode event happens
+- `/loop-ask`, `/loop-command`, and `/loop-shell` being configured but not reliably firing on their own
 
 The TUI loop is still intentionally session-bound: it runs while OpenCode is open and the current session emits status/idle events. For long-running background work after closing the terminal or OpenCode, use `opencode-loopd`.
 
+## v0.5.9 quick behavior guide
+
+OpenCode Loop has two triggers now:
+
+1. **Idle/status events** from OpenCode. These are used when OpenCode finishes a turn.
+2. **A real due timer** inside the plugin. This wakes delayed jobs even when no new OpenCode event happens.
+
+All TUI jobs are still idle-safe. If the timer expires while OpenCode is busy, the job does not interrupt the active turn. It waits and retries until the session becomes idle.
+
+Simple examples:
+
+```text
+/loop 0s continue the project
+```
+
+Run every time OpenCode becomes idle.
+
+```text
+/loop 1m --no-now continue the project
+```
+
+Wait 1 minute first, then run when OpenCode is idle.
+
+```text
+/loop-command 200m /compact
+```
+
+Run OpenCode compact about every 200 minutes, but only when OpenCode is idle.
+
+```text
+/loop-ask 1h did you run tests, tsc --noEmit, and build? If not, run them and fix errors.
+```
+
+Ask a recurring quality-control question every hour.
+
+```text
+/loop-shell 10m npm test
+```
+
+Run a shell command every 10 minutes when idle.
+
+```text
+/loop-goal --max-turns 5 --check "npm run build" --complete-when-checks-pass make the project build cleanly
+```
+
+Experimental persistent goal mode: keep working until the goal is complete, blocked, paused, cleared, or a safety limit is reached.
+
+> Note: OpenCode custom command markdown still creates a tiny assistant turn for slash commands. v0.5.9 keeps these command templates short and asks the model to reply `OK`. The actual loop scheduling is handled locally by the plugin. If that short turn briefly makes OpenCode look busy, the loop now recovers by refreshing stale busy status on the next due retry.
+
 ## Why this exists
 
 Claude Code users often rely on a loop-like workflow where the agent finishes one step, then immediately continues with the next step.

package/commands/loop-ask.md

@@ -2,4 +2,4 @@
 description: Schedule a recurring prompt/question. Usage: /loop-ask 1h did you run tests?
 ---
 
-Loop ask scheduler received. The OpenCode Loop plugin handles this recurring prompt locally. Do not explain this command, do not search the web, do not create scheduler files, and do not run extra shell commands because of this command.
+OpenCode Loop ask scheduler handled locally. Reply exactly: OK.

package/commands/loop-clear.md

@@ -1,5 +1,5 @@
 ---
-description: OpenCode Loop plugin command.
+description: Clear all OpenCode Loop jobs for this session.
 ---
 
-Loop command received. The OpenCode Loop plugin handles this command. Do not explain this command, do not search the web, do not create scheduler files, and do not run extra shell commands because of this command.
+OpenCode Loop clear command handled locally. Reply exactly: OK.

package/commands/loop-cmd.md

@@ -2,4 +2,4 @@
 description: Alias for /loop-command. Usage: /loop-cmd 200m /compact
 ---
 
-Loop command scheduler received. The OpenCode Loop plugin handles this slash-command schedule locally. Do not explain this command, do not search the web, do not create scheduler files, and do not run extra shell commands because of this command.
+OpenCode Loop command scheduler handled locally. Reply exactly: OK.

package/commands/loop-command.md

@@ -2,4 +2,4 @@
 description: Schedule an OpenCode slash command loop. Usage: /loop-command 200m /compact
 ---
 
-Loop command scheduler received. The OpenCode Loop plugin handles this slash-command schedule locally. Do not explain this command, do not search the web, do not create scheduler files, and do not run extra shell commands because of this command.
+OpenCode Loop command scheduler handled locally. Reply exactly: OK.

package/commands/loop-doctor.md

@@ -1,5 +1,5 @@
 ---
-description: OpenCode Loop plugin command.
+description: Run OpenCode Loop diagnostics.
 ---
 
-Loop command received. The OpenCode Loop plugin handles this command. Do not explain this command, do not search the web, do not create scheduler files, and do not run extra shell commands because of this command.
+OpenCode Loop doctor command handled locally. Reply exactly: OK.

package/commands/loop-export.md

@@ -1,5 +1,5 @@
 ---
-description: OpenCode Loop plugin command.
+description: Export OpenCode Loop state.
 ---
 
-Loop command received. The OpenCode Loop plugin handles this command. Do not explain this command, do not search the web, do not create scheduler files, and do not run extra shell commands because of this command.
+OpenCode Loop export command handled locally. Reply exactly: OK.

package/commands/loop-goal-blocked.md

@@ -1,5 +1,5 @@
 ---
-description: Manually mark the active experimental goal as blocked. Usage: /loop-goal-blocked reason
+description: Mark experimental goal blocked.
 ---
 
-Experimental goal blocked command received. The OpenCode Loop plugin handles this command locally.
+OpenCode Loop experimental goal blocked command handled locally. Reply exactly: OK.

package/commands/loop-goal-clear.md

@@ -1,5 +1,5 @@
 ---
-description: Clear experimental goal jobs. Usage: /loop-goal-clear
+description: Clear experimental goal jobs.
 ---
 
-Experimental goal clear command received. The OpenCode Loop plugin handles this command locally.
+OpenCode Loop experimental goal clear command handled locally. Reply exactly: OK.

package/commands/loop-goal-done.md

@@ -1,5 +1,5 @@
 ---
-description: Manually mark the active experimental goal as done. Usage: /loop-goal-done summary
+description: Mark experimental goal done.
 ---
 
-Experimental goal done command received. The OpenCode Loop plugin handles this command locally.
+OpenCode Loop experimental goal done command handled locally. Reply exactly: OK.

package/commands/loop-goal-pause.md

@@ -1,5 +1,5 @@
 ---
-description: Pause the active experimental goal. Usage: /loop-goal-pause
+description: Pause experimental goal jobs.
 ---
 
-Experimental goal pause command received. The OpenCode Loop plugin handles this command locally.
+OpenCode Loop experimental goal pause command handled locally. Reply exactly: OK.

package/commands/loop-goal-resume.md

@@ -1,5 +1,5 @@
 ---
-description: Resume the active experimental goal. Usage: /loop-goal-resume
+description: Resume experimental goal jobs.
 ---
 
-Experimental goal resume command received. The OpenCode Loop plugin handles this command locally.
+OpenCode Loop experimental goal resume command handled locally. Reply exactly: OK.

package/commands/loop-goal-status.md

@@ -1,5 +1,5 @@
 ---
-description: Show experimental goal status. Usage: /loop-goal-status
+description: Show experimental goal status.
 ---
 
-Experimental goal status command received. The OpenCode Loop plugin handles this command locally.
+OpenCode Loop experimental goal status command handled locally. Reply exactly: OK.

package/commands/loop-goal.md

@@ -2,4 +2,4 @@
 description: Experimental persistent goal mode. Usage: /loop-goal finish the feature and keep tests green
 ---
 
-Experimental goal command received. The OpenCode Loop plugin handles this goal locally. Do not explain this command, do not search the web, do not create scheduler files, and do not run extra shell commands because of this command.
+OpenCode Loop experimental goal command handled locally. Reply exactly: OK.

package/commands/loop-help.md

@@ -2,4 +2,4 @@
 description: Show OpenCode Loop help.
 ---
 
-Loop command received. The OpenCode Loop plugin handles this command. Do not explain this command, do not search the web, do not create scheduler files, and do not run extra shell commands because of this command.
+OpenCode Loop help command handled locally. Reply exactly: OK.

package/commands/loop-init.md

@@ -1,5 +1,5 @@
 ---
-description: OpenCode Loop plugin command.
+description: Create a starter progress.md file.
 ---
 
-Loop command received. The OpenCode Loop plugin handles this command. Do not explain this command, do not search the web, do not create scheduler files, and do not run extra shell commands because of this command.
+OpenCode Loop init command handled locally. Reply exactly: OK.

package/commands/loop-logs.md

@@ -2,4 +2,4 @@
 description: Show OpenCode Loop logs.
 ---
 
-Loop command received. The OpenCode Loop plugin handles this command. Do not explain this command, do not search the web, do not create scheduler files, and do not run extra shell commands because of this command.
+OpenCode Loop logs command handled locally. Reply exactly: OK.

package/commands/loop-now.md

@@ -1,5 +1,5 @@
 ---
-description: OpenCode Loop plugin command.
+description: Run due OpenCode Loop jobs now when idle.
 ---
 
-Loop command received. The OpenCode Loop plugin handles this command. Do not explain this command, do not search the web, do not create scheduler files, and do not run extra shell commands because of this command.
+OpenCode Loop run-now command handled locally. Reply exactly: OK.

package/commands/loop-pause.md

@@ -1,5 +1,5 @@
 ---
-description: OpenCode Loop plugin command.
+description: Pause OpenCode Loop jobs.
 ---
 
-Loop command received. The OpenCode Loop plugin handles this command. Do not explain this command, do not search the web, do not create scheduler files, and do not run extra shell commands because of this command.
+OpenCode Loop pause command handled locally. Reply exactly: OK.

package/commands/loop-resume.md

@@ -1,5 +1,5 @@
 ---
-description: OpenCode Loop plugin command.
+description: Resume OpenCode Loop jobs.
 ---
 
-Loop command received. The OpenCode Loop plugin handles this command. Do not explain this command, do not search the web, do not create scheduler files, and do not run extra shell commands because of this command.
+OpenCode Loop resume command handled locally. Reply exactly: OK.

package/commands/loop-shell.md

@@ -2,4 +2,4 @@
 description: Schedule a recurring shell command. Usage: /loop-shell 10m npm test
 ---
 
-Loop shell scheduler received. The OpenCode Loop plugin handles this shell schedule locally. Do not explain this command, do not search the web, do not create scheduler files, and do not run extra shell commands because of this command.
+OpenCode Loop shell scheduler handled locally. Reply exactly: OK.

package/commands/loop-status.md

@@ -2,4 +2,4 @@
 description: Show OpenCode Loop status.
 ---
 
-Loop command received. The OpenCode Loop plugin handles this command. Do not explain this command, do not search the web, do not create scheduler files, and do not run extra shell commands because of this command.
+OpenCode Loop status command handled locally. Reply exactly: OK.

package/commands/loop-stop.md

@@ -2,4 +2,4 @@
 description: Stop OpenCode Loop jobs.
 ---
 
-Loop command received. The OpenCode Loop plugin handles this command. Do not explain this command, do not search the web, do not create scheduler files, and do not run extra shell commands because of this command.
+OpenCode Loop stop command handled locally. Reply exactly: OK.

package/commands/loop.md

@@ -2,4 +2,4 @@
 description: Start an OpenCode auto-continue loop. Usage: /loop 5m <task>
 ---
 
-Loop command received. The OpenCode Loop plugin handles this command. Do not explain this command, do not search the web, do not create scheduler files, and do not run extra shell commands because of this command.
+OpenCode Loop local command handled. Reply exactly: OK.

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@bybrawe/opencode-loop",
-  "version": "0.5.7",
-  "description": "Claude Code/Codex style /loop and experimental goal mode for OpenCode: auto-continue, command loops, compact scheduling, verification, checkpoints, and safe autonomous coding goals.",
+  "version": "0.5.9",
+  "description": "Claude Code/Codex style /loop and experimental goal mode for OpenCode: reliable timers, stale-busy recovery, idle-safe loops, scheduled commands, compact scheduling, verification, checkpoints, and persistent coding goals.",
   "type": "module",
   "main": "src/index.js",
   "exports": {

package/src/index.js

@@ -6,6 +6,10 @@ const SERVICE = "opencode-loop"
 const STATE_DIR = ".opencode/opencode-loop"
 const DEFAULT_ACTIVE_GUARD_MS = 60_000
 const IDLE_DEBOUNCE_MS = 1_200
+const BUSY_RETRY_MS = 5_000
+const SESSION_STATUS_CACHE_MS = 1_500
+const MIN_DUE_TIMER_MS = 250
+const MAX_DUE_TIMER_MS = 2_147_000_000
 const MAX_SCAN_FILES = 200
 const MAX_SCAN_BYTES = 2_000_000
 const GOAL_REPORT_DIR = "goals"
@@ -14,7 +18,9 @@ const GOAL_PROMPT_PREFIX = "EXPERIMENTAL OPENCODE GOAL MODE ITERATION"
 const activeRuns = new Map()
 const handledCommands = new Map()
 const idleTimers = new Map()
+const dueTimers = new Map()
 const sessionStatuses = new Map()
+const sessionStatusSeenAt = new Map()
 
 const DEFAULT_PROGRESS_MD = `# Progress
 
@@ -385,16 +391,22 @@ async function executeTuiCommand(client, command) {
 
 function compactTuiCommandName(command = "compact") {
   const normalized = String(command || "compact").replace(/^\/+/, "").trim().toLowerCase()
-  if (normalized === "compact" || normalized === "summarize") return "session.compact"
+  if (normalized === "compact" || normalized === "summarize") return "session_compact"
   return undefined
 }
 
 async function compactSession(client, sessionID) {
-  try {
-    await executeTuiCommand(client, "session.compact")
-    return true
-  } catch (error) {
-    await log(client, "warn", "tui session.compact failed", { error: sdkErrorMessage(error) })
+  // OpenCode's TUI API accepts legacy keybind aliases (session_compact) in
+  // current builds, while some older docs/examples mention the event value
+  // (session.compact). Try the alias first, then the event value, then the
+  // session summarize endpoint as a last resort.
+  for (const command of ["session_compact", "session.compact"]) {
+    try {
+      await executeTuiCommand(client, command)
+      return true
+    } catch (error) {
+      await log(client, "warn", `tui ${command} failed`, { error: sdkErrorMessage(error) })
+    }
   }
   try {
     await sdkCall(client.session.summarize.bind(client.session), { sessionID }, { path: { id: sessionID }, body: {} })
@@ -730,12 +742,16 @@ function updateSessionStatusFromEvent(event) {
   if (typeof sessionID !== "string") return undefined
   if (event?.type === "session.idle") {
     sessionStatuses.set(sessionID, "idle")
+    sessionStatusSeenAt.set(sessionID, now())
     return { sessionID, idle: true }
   }
   if (event?.type === "session.status") {
     const status = event?.properties?.status
     const type = status && typeof status === "object" ? status.type : undefined
-    if (typeof type === "string") sessionStatuses.set(sessionID, type)
+    if (typeof type === "string") {
+      sessionStatuses.set(sessionID, type)
+      sessionStatusSeenAt.set(sessionID, now())
+    }
     return { sessionID, idle: type === "idle" }
   }
   return undefined
@@ -743,18 +759,31 @@ function updateSessionStatusFromEvent(event) {
 
 async function sessionStatusType(client, sessionID) {
   const cached = sessionStatuses.get(sessionID)
-  if (cached && cached !== "idle") return cached
+  const seenAt = sessionStatusSeenAt.get(sessionID) || 0
+
+  // Idle is safe to trust until OpenCode tells us otherwise. Busy/retry is only
+  // trusted briefly: OpenCode custom commands such as /loop-status create their
+  // own short assistant turn, and some TUI builds do not always emit the final
+  // idle event after that turn. If we cache busy forever, due loop work can get
+  // stuck at "due in every idle" until the user types another command.
+  if (cached === "idle") return cached
+  if (cached && now() - seenAt < SESSION_STATUS_CACHE_MS) return cached
+
   try {
     const data = await sdkCall(client.session.status.bind(client.session), {}, undefined)
     const status = data?.[sessionID]
     const type = status && typeof status === "object" ? status.type : undefined
     if (typeof type === "string") {
       sessionStatuses.set(sessionID, type)
+      sessionStatusSeenAt.set(sessionID, now())
       return type
     }
   } catch {}
-  sessionStatuses.set(sessionID, "idle")
-  return "idle"
+
+  const fallback = activeRuns.has(sessionID) ? "busy" : "idle"
+  sessionStatuses.set(sessionID, fallback)
+  sessionStatusSeenAt.set(sessionID, now())
+  return fallback
 }
 
 async function sessionIsIdle(client, sessionID) {
@@ -768,9 +797,15 @@ function scheduleIdleWork(directory, client, sessionID) {
     idleTimers.delete(sessionID)
     Promise.resolve()
       .then(async () => {
-        if (!await sessionIsIdle(client, sessionID)) return
+        if (!await sessionIsIdle(client, sessionID)) {
+          await scheduleDueWork(directory, client, sessionID, BUSY_RETRY_MS)
+          return
+        }
         await finalizeActiveRun(directory, client, sessionID)
-        if (!await sessionIsIdle(client, sessionID)) return
+        if (!await sessionIsIdle(client, sessionID)) {
+          await scheduleDueWork(directory, client, sessionID, BUSY_RETRY_MS)
+          return
+        }
         await maybeRunDueJobs(directory, client, sessionID)
       })
       .catch((error) => {
@@ -781,6 +816,61 @@ function scheduleIdleWork(directory, client, sessionID) {
   idleTimers.set(sessionID, timer)
 }
 
+function jobDueAt(job, current = now()) {
+  if (isGoalJob(job) && ["completed", "blocked", "cleared"].includes(job.goalStatus)) return Infinity
+  if (!job.enabled || job.paused) return Infinity
+  if (job.maxRuns > 0 && (job.runCount || 0) >= job.maxRuns) return Infinity
+  if (job.watchPaths?.length) return Infinity
+  const created = Date.parse(job.createdAt || new Date().toISOString())
+  if (job.maxRuntimeMs > 0 && current - created >= job.maxRuntimeMs) return current
+  if (job.intervalMs === 0) return current
+  if (!job.lastRunAt) return current
+  return job.lastRunAt + (job.intervalMs || 0)
+}
+
+function nextDueDelay(state) {
+  const current = now()
+  let soonest = Infinity
+  for (const job of state.jobs || []) soonest = Math.min(soonest, jobDueAt(job, current))
+  if (!Number.isFinite(soonest)) return Infinity
+  return Math.max(0, soonest - current)
+}
+
+async function scheduleDueWork(directory, client, sessionID, minDelayMs = 0) {
+  const previous = dueTimers.get(sessionID)
+  if (previous) clearTimeout(previous)
+
+  const state = await readState(directory, sessionID)
+  const delay = nextDueDelay(state)
+  if (!Number.isFinite(delay)) {
+    dueTimers.delete(sessionID)
+    return
+  }
+
+  const wait = Math.min(Math.max(delay, minDelayMs, MIN_DUE_TIMER_MS), MAX_DUE_TIMER_MS)
+  const timer = setTimeout(() => {
+    dueTimers.delete(sessionID)
+    Promise.resolve()
+      .then(async () => {
+        if (!await sessionIsIdle(client, sessionID)) {
+          await scheduleDueWork(directory, client, sessionID, BUSY_RETRY_MS)
+          return
+        }
+        await finalizeActiveRun(directory, client, sessionID)
+        if (!await sessionIsIdle(client, sessionID)) {
+          await scheduleDueWork(directory, client, sessionID, BUSY_RETRY_MS)
+          return
+        }
+        await maybeRunDueJobs(directory, client, sessionID)
+      })
+      .catch((error) => {
+        toast(client, `Loop due timer failed: ${sdkErrorMessage(error)}`, "error").catch(() => {})
+        appendLoopLog(directory, "due-timer-error", { sessionID, error: sdkErrorMessage(error) }).catch(() => {})
+      })
+  }, wait)
+  dueTimers.set(sessionID, timer)
+}
+
 function dueJobs(state, force = false) {
   const current = now()
   return (state.jobs || []).filter((job) => {
@@ -992,6 +1082,7 @@ async function finalizeActiveRun(directory, client, sessionID) {
   await writeState(directory, sessionID, state)
   if (isGoalJob(job)) await writeGoalReport(directory, sessionID, job)
   await createCheckpoint(directory, sessionID, job, client)
+  await scheduleDueWork(directory, client, sessionID)
 }
 
 async function fireAction(directory, client, sessionID, job) {
@@ -1010,7 +1101,7 @@ async function fireAction(directory, client, sessionID, job) {
     }
     const tuiCommand = compactTuiCommandName(command)
     if (tuiCommand) {
-      await executeTuiCommand(client, tuiCommand)
+      await compactSession(client, sessionID)
       return { startsAssistantTurn: true }
     }
     await sdkCall(client.session.command.bind(client.session), { sessionID, command, arguments: argumentsText }, { path: { id: sessionID }, body: { command, arguments: argumentsText } })
@@ -1039,19 +1130,29 @@ ${prompt}` }] } })
 }
 
 async function maybeRunDueJobs(directory, client, sessionID, options = {}) {
+  const reschedule = async (minDelayMs = 0) => { await scheduleDueWork(directory, client, sessionID, minDelayMs) }
+
   if (!await sessionIsIdle(client, sessionID)) {
-    if (options.force) await toast(client, "Loop queued: session is busy; it will run on the next idle event.", "info")
+    if (options.force) await toast(client, "Loop queued: session is busy; it will run on the next idle check.", "info")
+    await reschedule(BUSY_RETRY_MS)
     return
   }
   const active = activeRuns.get(sessionID)
-  if (active && active.job?.noOverlap !== false && now() - active.startedAt < (active.job?.timeoutMs || DEFAULT_ACTIVE_GUARD_MS)) return
+  if (active && active.job?.noOverlap !== false && now() - active.startedAt < (active.job?.timeoutMs || DEFAULT_ACTIVE_GUARD_MS)) {
+    await reschedule(BUSY_RETRY_MS)
+    return
+  }
 
   const state = await readState(directory, sessionID)
   for (const job of state.jobs || []) {
     if (job.watchPaths?.length && !job.paused && job.enabled && await watchChanged(directory, job)) job.lastRunAt = 0
   }
   let due = dueJobs(state, options.force)
-  if (!due.length) { await writeState(directory, sessionID, state); return }
+  if (!due.length) {
+    await writeState(directory, sessionID, state)
+    await reschedule()
+    return
+  }
   let job = due[0]
 
   if (job.maxRuntimeMs > 0 && now() - Date.parse(job.createdAt || new Date().toISOString()) >= job.maxRuntimeMs) {
@@ -1060,6 +1161,7 @@ async function maybeRunDueJobs(directory, client, sessionID, options = {}) {
     await notifyJob(directory, job, "max_runtime_reached")
     await toast(client, `Loop stopped by --max-runtime: ${job.name || job.id}`, "success")
     await appendLoopLog(directory, "max-runtime", { sessionID, job: job.name || job.id })
+    await reschedule()
     return
   }
   if (job.stopFile && await pathExists(path.resolve(directory, job.stopFile))) {
@@ -1067,6 +1169,7 @@ async function maybeRunDueJobs(directory, client, sessionID, options = {}) {
     await writeState(directory, sessionID, state)
     await notifyJob(directory, job, "stop_file")
     await toast(client, "Loop stopped by --stop-file: " + job.stopFile, "success")
+    await reschedule()
     return
   }
   if (await untilReached(directory, job)) {
@@ -1074,6 +1177,7 @@ async function maybeRunDueJobs(directory, client, sessionID, options = {}) {
     await writeState(directory, sessionID, state)
     await notifyJob(directory, job, "until_reached")
     await toast(client, `Loop stopped by --until: ${job.until}`, "success")
+    await reschedule()
     return
   }
 
@@ -1083,6 +1187,7 @@ async function maybeRunDueJobs(directory, client, sessionID, options = {}) {
       await writeState(directory, sessionID, state)
       await notifyJob(directory, job, "preflight_blocked")
       await toast(client, "Preflight blocked in safe mode and loop paused: " + job.preflightCommand, "error")
+      await reschedule()
       return
     }
     const preflight = await runShellCommand(job.preflightCommand, directory, job.timeoutMs || 300_000)
@@ -1095,6 +1200,7 @@ async function maybeRunDueJobs(directory, client, sessionID, options = {}) {
       await writeState(directory, sessionID, state)
       await notifyJob(directory, job, "preflight_failed")
       await toast(client, "Preflight failed and loop paused: " + job.preflightCommand, "warning")
+      await reschedule()
       return
     }
   }
@@ -1118,16 +1224,18 @@ async function maybeRunDueJobs(directory, client, sessionID, options = {}) {
       const fresh = await readState(directory, sessionID)
       fresh.jobs = (fresh.jobs || []).filter((candidate) => candidate.enabled !== false || isGoalJob(candidate))
       await writeState(directory, sessionID, fresh)
+      await reschedule()
+      return
     }
-    if (result.startsAssistantTurn) {
-      let timer
-      if (job.timeoutMs > 0) timer = setTimeout(() => { fireSdk(client, "session.abort", client.session.abort.bind(client.session), { sessionID }, { path: { id: sessionID }, body: {} }); toast(client, `Loop timeout fired: ${job.name || job.id}`, "warning").catch(() => {}) }, job.timeoutMs)
-      activeRuns.set(sessionID, { jobId: job.id, job, startedAt: now(), timer })
-    }
+    let timer
+    if (job.timeoutMs > 0) timer = setTimeout(() => { fireSdk(client, "session.abort", client.session.abort.bind(client.session), { sessionID }, { path: { id: sessionID }, body: {} }); toast(client, `Loop timeout fired: ${job.name || job.id}`, "warning").catch(() => {}) }, job.timeoutMs)
+    activeRuns.set(sessionID, { jobId: job.id, job, startedAt: now(), timer })
+    await reschedule(BUSY_RETRY_MS)
   } catch (error) {
     clearActiveRun(sessionID)
     await toast(client, `Loop job failed: ${error instanceof Error ? error.message : String(error)}`, "error")
     await appendLoopLog(directory, "error", { sessionID, job: job.name || job.id, error: error instanceof Error ? error.message : String(error) })
+    await reschedule(BUSY_RETRY_MS)
   }
 }
 
@@ -1171,6 +1279,7 @@ async function addLoop(directory, client, sessionID, args, defaults = {}) {
 
   state.jobs.push(parsed.job)
   await writeState(directory, sessionID, state)
+  await scheduleDueWork(directory, client, sessionID)
   if (parsed.job.immediate) scheduleIdleWork(directory, client, sessionID)
   await toast(client, `${replaced ? "Loop replaced" : "Loop added"}: ${jobLabel(parsed.job)}`, "success")
   await appendLoopLog(directory, replaced ? "replace" : "add", { sessionID, job: parsed.job.name || parsed.job.id, label: jobLabel(parsed.job) })
@@ -1178,11 +1287,18 @@ async function addLoop(directory, client, sessionID, args, defaults = {}) {
 
 async function stopLoop(directory, client, sessionID, args) {
   const target = String(args || "").trim()
-  if (!target || target.toLowerCase() === "all") { await removeState(directory, sessionID); clearActiveRun(sessionID); await toast(client, "All loops stopped for this session.", "success"); return }
+  if (!target || target.toLowerCase() === "all") {
+    await removeState(directory, sessionID)
+    clearActiveRun(sessionID)
+    const due = dueTimers.get(sessionID); if (due) clearTimeout(due); dueTimers.delete(sessionID)
+    await toast(client, "All loops stopped for this session.", "success")
+    return
+  }
   const state = await readState(directory, sessionID)
   const before = state.jobs.length
   state.jobs = state.jobs.filter((job, index) => !matchJob(job, target, index))
   await writeState(directory, sessionID, state)
+  await scheduleDueWork(directory, client, sessionID)
   await toast(client, `Stopped ${before - state.jobs.length} loop(s).`, "success")
 }
 
@@ -1192,6 +1308,7 @@ async function updateJobState(directory, client, sessionID, args, updater, messa
   let count = 0
   state.jobs = (state.jobs || []).map((job, index) => matchJob(job, target, index) ? (count++, updater(job)) : job)
   await writeState(directory, sessionID, state)
+  await scheduleDueWork(directory, client, sessionID)
   await toast(client, `${message}: ${count} loop(s).`, count ? "success" : "warning")
 }
 
@@ -1215,6 +1332,7 @@ async function pauseGoal(directory, client, sessionID, args) {
   let count = 0
   state.jobs = (state.jobs || []).map((job, index) => isGoalJob(job) && matchJob(job, target, index) ? (count++, { ...job, paused: true }) : job)
   await writeState(directory, sessionID, state)
+  await scheduleDueWork(directory, client, sessionID)
   await toast(client, `Paused ${count} experimental goal(s).`, count ? "success" : "warning")
 }
 
@@ -1229,7 +1347,10 @@ async function resumeGoal(directory, client, sessionID, args) {
   })
   await writeState(directory, sessionID, state)
   await toast(client, `Resumed ${count} experimental goal(s).`, count ? "success" : "warning")
-  if (count) scheduleIdleWork(directory, client, sessionID)
+  if (count) {
+    await scheduleDueWork(directory, client, sessionID)
+    scheduleIdleWork(directory, client, sessionID)
+  }
 }
 
 async function clearGoal(directory, client, sessionID, args) {
@@ -1238,6 +1359,7 @@ async function clearGoal(directory, client, sessionID, args) {
   const before = state.jobs.length
   state.jobs = (state.jobs || []).filter((job, index) => !isGoalJob(job) || (target && !matchJob(job, target, index)))
   await writeState(directory, sessionID, state)
+  await scheduleDueWork(directory, client, sessionID)
   await toast(client, `Cleared ${before - state.jobs.length} experimental goal(s).`, before !== state.jobs.length ? "success" : "warning")
 }