# Send Autonomous Task Runner to your agent
Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually.
## Fast path
- Download the package from Yavira.
- Extract it into a folder your agent can access.
- Paste one of the prompts below and point your agent at the extracted folder.
## Suggested prompts
### New install

```text
I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Then review README.md for any prerequisites, environment setup, or post-install checks. Tell me what you changed and call out any manual steps you could not complete.
```
### Upgrade existing

```text
I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Then review README.md for any prerequisites, environment setup, or post-install checks. Summarize what changed and any follow-up checks I should run.
```
## Machine-readable fields
```json
{
  "schemaVersion": "1.0",
  "item": {
    "slug": "autonomous-task-runner",
    "name": "Autonomous Task Runner",
    "source": "tencent",
    "type": "skill",
    "category": "AI 智能",
    "sourceUrl": "https://clawhub.ai/chunhualiao/autonomous-task-runner",
    "canonicalUrl": "https://clawhub.ai/chunhualiao/autonomous-task-runner",
    "targetPlatform": "OpenClaw"
  },
  "install": {
    "downloadUrl": "/downloads/autonomous-task-runner",
    "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=autonomous-task-runner",
    "sourcePlatform": "tencent",
    "targetPlatform": "OpenClaw",
    "packageFormat": "ZIP package",
    "primaryDoc": "SKILL.md",
    "includedAssets": [
      "CHANGELOG.md",
      "README.md",
      "SKILL.md",
      "STATUS.json",
      "references/queue-schema.md",
      "references/task-types.md"
    ],
    "downloadMode": "redirect",
    "sourceHealth": {
      "source": "tencent",
      "slug": "autonomous-task-runner",
      "status": "healthy",
      "reason": "direct_download_ok",
      "recommendedAction": "download",
      "checkedAt": "2026-04-30T00:29:37.360Z",
      "expiresAt": "2026-05-07T00:29:37.360Z",
      "httpStatus": 200,
      "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=autonomous-task-runner",
      "contentType": "application/zip",
      "probeMethod": "head",
      "details": {
        "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=autonomous-task-runner",
        "contentDisposition": "attachment; filename=\"autonomous-task-runner-2.1.0.zip\"",
        "redirectLocation": null,
        "bodySnippet": null,
        "slug": "autonomous-task-runner"
      },
      "scope": "item",
      "summary": "Item download looks usable.",
      "detail": "Yavira can redirect you to the upstream package for this item.",
      "primaryActionLabel": "Download for OpenClaw",
      "primaryActionHref": "/downloads/autonomous-task-runner"
    },
    "validation": {
      "installChecklist": [
        "Use the Yavira download entry.",
        "Review SKILL.md after the package is downloaded.",
        "Confirm the extracted package contains the expected setup assets."
      ],
      "postInstallChecks": [
        "Confirm the extracted package includes the expected docs or setup files.",
        "Validate the skill or prompts are available in your target agent workspace.",
        "Capture any manual follow-up steps the agent could not complete."
      ]
    }
  },
  "links": {
    "detailUrl": "https://openagent3.xyz/skills/autonomous-task-runner",
    "downloadUrl": "https://openagent3.xyz/downloads/autonomous-task-runner",
    "agentUrl": "https://openagent3.xyz/skills/autonomous-task-runner/agent",
    "manifestUrl": "https://openagent3.xyz/skills/autonomous-task-runner/agent.json",
    "briefUrl": "https://openagent3.xyz/skills/autonomous-task-runner/agent.md"
  }
}
```
## Documentation

### Task Runner Skill

A persistent, daemon-style task queue. Users add tasks at any time. A dispatcher runs on every
heartbeat to check the queue and execute pending work via subagents. Tasks accumulate, complete,
and are archived — the queue itself never closes.

### Two Operating Modes

This skill has two distinct modes with different triggers and behaviors:

ModeTriggerPurposeINTAKEUser message containing task intentParse message → add tasks to queue → confirm → immediately run DISPATCHERDISPATCHERAfter INTAKE (primary) · Heartbeat/cron (backup)Read queue → dispatch pending tasks → report completions

Both modes read and write the same persistent queue file.

### Mode 1: INTAKE (user message)

Activate INTAKE mode when the user's message matches any of the following patterns:

PatternExamplesExplicit task add"add task", "add these tasks", "task:", "new task"Delegation"do this for me", "do these for me", "handle these", "can you do X"Framing"I need you to", "help me with", "I need", "I want you to"List framing"task list", "my tasks", "queue these", "work on these"Control commands"skip T-03", "retry T-02", "mark T-01 done", "cancel T-04"Status check"show tasks", "task status", "what's in the queue", "what are my pending tasks"Compound askAny message with 2+ distinct action items (bullets, numbers, "and also", "then")

Do NOT activate INTAKE for:

Pure single-question lookups answered in one sentence ("what time is it?")
Scheduling-only requests with no actual task ("remind me in 20 min")
Single web search requests ("google X")
The heartbeat systemEvent (that's DISPATCHER mode)

### Mode 2: DISPATCHER (inline after INTAKE, heartbeat, or cron)

Activate DISPATCHER mode when triggered by:

Immediately after INTAKE — runs in the same turn, right after tasks are queued (primary path)
HEARTBEAT.md check during a heartbeat poll (backup: catches retries and completions)
systemEvent: "TASK_RUNNER_DISPATCH: check queue and run pending tasks" (backup)
Any scheduled/cron trigger registered for task-runner (backup)

### Configuration

VariableLocationDefaultDescriptionTASK_RUNNER_DIRTOOLS.md~/.openclaw/tasks/Directory for queue file and deliverablesTASK_RUNNER_MAX_CONCURRENTTOOLS.md2Max tasks running simultaneouslyTASK_RUNNER_MAX_RETRIESTOOLS.md or env3Max retry attempts before marking blockedTASK_RUNNER_ARCHIVE_DAYSTOOLS.md7Days after which done/blocked tasks are archived

How to configure — add to TOOLS.md:

## Task Runner
TASK_RUNNER_DIR=~/.openclaw/tasks/
TASK_RUNNER_MAX_CONCURRENT=2
TASK_RUNNER_MAX_RETRIES=3
TASK_RUNNER_ARCHIVE_DAYS=7

Queue file path: ${TASK_RUNNER_DIR}/task-queue.json
(single persistent file, NOT dated — accumulates all tasks over time)

### A3 — Outputs

OutputPath / ChannelDescriptionQueue file${TASK_RUNNER_DIR}/task-queue.jsonSingle persistent queue; all tasksPer-task completion messageChat notificationSent immediately when a task finishes (done or blocked)Deliverable filesTask-specific pathsFiles produced by tasks (when applicable)INTAKE confirmationChatSent after adding tasks to queue

### Mode 1: INTAKE — Step-by-Step

Goal: Convert user message into structured task objects, append to queue, confirm.

### Step 0 — First Run Setup (auto-configure on first use)

Run this check before anything else, every INTAKE invocation:

CHECK whether ${TASK_RUNNER_DIR}/task-queue.json exists
IF file does NOT exist:
  → This is the first run. Auto-configure everything silently before proceeding.

  [1] Create directory:
      exec: mkdir -p ${TASK_RUNNER_DIR}

  [2] Initialize queue file:
      WRITE ${TASK_RUNNER_DIR}/task-queue.json with default structure:
      { "lastId": null, "tasks": [], "archivedCount": 0 }

  [3] Register heartbeat entry:
      READ HEARTBEAT.md (create it if missing)
      IF "Task Runner Dispatcher" is NOT already in the file:
        APPEND the following block (with a blank line before it):

        ## Task Runner Dispatcher
        Every heartbeat: check ${TASK_RUNNER_DIR}/task-queue.json
        - If pending or running tasks exist → run DISPATCHER mode (task-runner skill)
        - If nothing pending → HEARTBEAT_OK (skip)

      WRITE the updated HEARTBEAT.md

  [4] Register backup cron job:
      CALL cron tool with:
        action: "add"
        job:
          name: "Task Runner Dispatcher"
          schedule: { kind: "every", everyMs: 900000 }
          payload: { kind: "systemEvent", text: "TASK_RUNNER_DISPATCH: check queue and run pending tasks" }
          sessionTarget: "main"
          enabled: true

  [5] Notify user:
      "⚙️ Task Runner initialized.
       Heartbeat dispatcher registered in HEARTBEAT.md.
       Backup cron job registered (runs every 15 minutes).
       Your tasks will execute automatically."

  → THEN continue with normal INTAKE steps below.

IF file already exists:
  → Skip Step 0 entirely. Proceed directly to Step 1.

Idempotency rule: Step 0 only fires on true first run (queue file absent).
It will never double-register the heartbeat entry or create duplicate cron jobs.

### Step 1 — Load queue

READ ${TASK_RUNNER_DIR}/task-queue.json
IF file does not exist:
  Initialize with default structure (see references/queue-schema.md)
  Set lastId = null

### Step 2 — Parse tasks from message

Split user message into individual tasks using these cues:

Numbered lists (1., 2., 3.)
Bulleted lists (-, *, •)
Explicit separators ("first", "also", "and then", "next")
Compound sentences with multiple imperatives
Single task: entire message is one task

### Step 3 — Assign IDs

Continue from lastId in the queue file:

If lastId = "T-05", next task is T-06
If lastId = null, start at T-01
Format: T-NN (zero-padded, minimum 2 digits; expand to 3 when N > 99)

### Step 4 — Build task objects

For each parsed task, create a JSON object (schema in references/queue-schema.md):

Set id, description, goal, status = "pending", added_at
Set retries = 0, maxRetries from config
Leave execution fields null

### Step 5 — Append to queue and save

APPEND new task objects to queue.tasks[]
UPDATE queue.lastId to the last assigned ID
WRITE updated queue file to disk

### Step 6 — Confirm to user

Added T-06: [description]. Starting now...

For multiple tasks:

📋 Added 3 tasks to queue:
• T-06: [description]
• T-07: [description]
• T-08: [description]
Starting dispatcher now...

Then immediately run DISPATCHER mode (Steps 1–5 below) in the same turn.
Do not exit and wait for the next heartbeat. Tasks must start executing immediately.
The heartbeat/cron dispatcher is a backup for retries and completion checks — not the primary execution path.

### Step 7 — Handle control commands

CommandActionskip T-NNSet status = "skipped"; save; confirmretry T-NNReset status = "pending", retries = 0; save; confirmcancel T-NNSet status = "skipped", blocked_reason = "cancelled by user"; save; confirmmark T-NN doneSet status = "done", completed_at = now; save; confirmshow tasks / task statusRead queue; render status table (see A5 templates)

### Mode 2: DISPATCHER — Step-by-Step

Goal: Check queue, dispatch pending tasks, track running tasks, report completions.

### Step 1 — Load queue

READ ${TASK_RUNNER_DIR}/task-queue.json
IF file does not exist OR tasks array is empty:
  → HEARTBEAT_OK (silent, nothing to do)
  → EXIT

### Step 2 — Check for work

pending_tasks = tasks where status = "pending"
running_tasks = tasks where status = "running"

IF pending_tasks is empty AND running_tasks is empty:
  → HEARTBEAT_OK (silent)
  → EXIT

### Step 3 — Check running tasks for completion

For each task with status = "running":

IF subagent_session is set:
  CHECK subagent session status

  IF session is DONE:
    READ deliverable from session output
    RUN verification (see references/verification-guide.md)
    IF verification passes:
      SET status = "done"
      SET deliverable, deliverable_path, completed_at
      NOTIFY user: ✅ T-NN done — [summary]
    ELSE (verification failed):
      TREAT as failure (see retry logic below)

  IF session is FAILED or ERROR:
    IF retries < maxRetries:
      INCREMENT retries
      ADD to strategies_tried
      SET status = "pending"  ← will be re-dispatched this cycle
    ELSE:
      SET status = "blocked"
      SET blocked_reason, user_action_required, completed_at
      NOTIFY user: 🚫 T-NN blocked — [reason + unblock steps]

  IF session is STILL RUNNING:
    Leave as-is (will check again next heartbeat)

### Step 4 — Dispatch pending tasks

currently_running = count of tasks with status = "running"
slots_available = maxConcurrent - currently_running

FOR EACH pending task (in order of added_at), up to slots_available:
  PICK execution strategy (see references/task-types.md)
  SPAWN subagent with task description and strategy
  SET status = "running"
  SET subagent_session = spawned session ID
  SET started_at = now

Subagent instructions template:

You are executing task [T-NN] for the task-runner skill.

Task: [description]
Goal: [goal]
Type: [task_type]
Strategy: [selected strategy from task-types.md]

Execute the task. When complete:
1. Report the result clearly
2. Note any deliverable file path if a file was created
3. If blocked, explain exactly why and what the user needs to do

Do not start any other tasks. Focus only on this one.

### Step 5 — Save and exit

WRITE updated queue file (status changes, subagent_session IDs)

If any notifications were sent (done/blocked), this is an active heartbeat response.
If only silent dispatching occurred, this is still a heartbeat response (not HEARTBEAT_OK).
Only return HEARTBEAT_OK when there was truly nothing to do (no pending, no running tasks).

### INTAKE confirmation (single task)

Added T-06: [description]. Queue now has N pending tasks.

### INTAKE confirmation (multiple tasks)

📋 Added N tasks to queue:
• T-06: [description]
• T-07: [description]

Starting now...

### Task status table (on demand)

📋 Task Queue — [N total, N pending, N running, N done, N blocked]

ID    Status      Description
T-01  ✅ done      [description] → [deliverable summary]
T-02  🔄 running   [description] (started [time ago])
T-03  ⏳ pending   [description]
T-04  🚫 blocked   [description] — [blocked_reason short]
T-05  ⏭️ skipped   [description]

### Task done notification

✅ T-NN done — [one-sentence summary of what was accomplished]
[deliverable: link or file path, if applicable]

### Task blocked notification

🚫 T-NN blocked after [N] attempts

What was tried:
- [Strategy 1]: [result]
- [Strategy 2]: [result]

Why it's blocked:
[Clear plain-English explanation]

To unblock:
1. [Concrete step #1]
2. [Concrete step #2 if needed]

Reply "retry T-NN" once ready.

### Task skipped

⏭️ T-NN skipped — as requested.

### A6 — Heartbeat Integration

Heartbeat and cron setup is automatic. Step 0 of INTAKE mode handles this on first use —
no manual configuration required.

### Role of heartbeat/cron (backup only)

Tasks are dispatched immediately after INTAKE — heartbeat and cron are backups only.

The backup dispatcher handles:

Retry dispatch: tasks that failed and were reset to pending
Completion checks: polling running subagent sessions for done/blocked status
Recovery: tasks that were pending when no user message triggered INTAKE

Users should never need to wait for a heartbeat for a freshly added task.

### What gets configured automatically

HEARTBEAT.md entry (injected on first INTAKE):

## Task Runner Dispatcher
Every heartbeat: check ${TASK_RUNNER_DIR}/task-queue.json
- If pending or running tasks exist → run DISPATCHER mode (task-runner skill)
- If nothing pending → HEARTBEAT_OK (skip)

Backup cron job (registered on first INTAKE):

every 15 min → systemEvent: "TASK_RUNNER_DISPATCH: check queue and run pending tasks"
sessionTarget: main

### Manual setup (if needed)

If for any reason auto-setup did not run (e.g., queue file was pre-created externally),
delete ${TASK_RUNNER_DIR}/task-queue.json and send any task — Step 0 will fire.

### INTAKE mode succeeds when:

All tasks from user message parsed and assigned IDs
Tasks appended to queue file (file saved to disk)
Confirmation sent to user with task IDs and count
DISPATCHER mode triggered immediately in the same turn
Subagents spawned for pending tasks before INTAKE turn ends

### DISPATCHER mode succeeds when:

Queue file read without error
All running tasks checked for completion (done/blocked notifications sent as needed)
Pending tasks dispatched up to maxConcurrent slots
Queue file saved with updated states
User notified for every task that reached a terminal state this cycle

### Ongoing system health:

Queue file is never corrupted (always valid JSON)
Tasks older than archiveDays days with terminal status are archived/removed
lastId always increments (no ID reuse)
maxRetries respected before any task is marked blocked

### Edge Cases

SituationBehaviorQueue file missing (first run)Run Step 0 auto-setup: create dir, init queue, register heartbeat + cron; notify userQueue file missing (manually deleted)Step 0 re-fires: re-initializes queue; does NOT re-register heartbeat/cron (idempotent check)Queue file corrupt/invalid JSONLog error, notify user, do not overwrite; ask user to inspectTask description is ambiguousAssign unknown type; dispatcher will attempt classification + fallbackmaxConcurrent already reachedDispatcher skips dispatching; checks again next heartbeatUser adds task while dispatcher is runningRace-safe: dispatcher reads, processes, writes atomically per cycleTask depends on another task's outputSet blocked_reason = "depends on T-NN-1 which is pending/blocked"User says "retry T-NN"Reset to pending, retries = 0, strategies_tried = []All tasks blockedNotify user: "All tasks are blocked. Review unblock instructions above."20+ tasks added at onceDispatcher dispatches in batches of maxConcurrent; all tasks eventually runSubagent session ID lostMark task as pending again; will re-dispatch next cycleArchive: done tasks > archiveDays oldMove to ${TASK_RUNNER_DIR}/archive/YYYY-MM.json; remove from main queue

### A8 — File Organization

${TASK_RUNNER_DIR}/
  task-queue.json          ← single persistent queue (all active tasks)
  archive/
    2026-01.json           ← archived tasks (done/blocked, older than archiveDays)
    2026-02.json

Queue file schema is documented in references/queue-schema.md.

### References

references/queue-schema.md — Queue JSON format (complete field reference)
references/task-types.md — Task type catalog and strategy selection
references/verification-guide.md — Verification logic per task type
tests/test-triggers.json — Trigger test cases (positive and negative)
## Trust
- Source: tencent
- Verification: Indexed source record
- Publisher: chunhualiao
- Version: 2.1.0
## Source health
- Status: healthy
- Item download looks usable.
- Yavira can redirect you to the upstream package for this item.
- Health scope: item
- Reason: direct_download_ok
- Checked at: 2026-04-30T00:29:37.360Z
- Expires at: 2026-05-07T00:29:37.360Z
- Recommended action: Download for OpenClaw
## Links
- [Detail page](https://openagent3.xyz/skills/autonomous-task-runner)
- [Send to Agent page](https://openagent3.xyz/skills/autonomous-task-runner/agent)
- [JSON manifest](https://openagent3.xyz/skills/autonomous-task-runner/agent.json)
- [Markdown brief](https://openagent3.xyz/skills/autonomous-task-runner/agent.md)
- [Download page](https://openagent3.xyz/downloads/autonomous-task-runner)