Skip to content

LeoYeAI/agent-master-skills / skills/autonomous-task-runner

task-runner

Verified

Persistent task queue system. Users add tasks at any time via natural language; tasks are stored in a single persistent queue file and executed asynchronously via subagents. A heartbeat/cron dispatcher wakes periodically to check pending tasks, spawn workers, and report completions. The system never "finishes" — it always remains ready for the next task.

FileKindSize
CHANGELOG.mdmarkdown2.8 KB
README.mdmarkdown6.4 KB
SKILL.mdskill md16.1 KB
STATUS.jsonconfig653 B
_meta.jsonconfig301 B
references/queue-schema.mdmarkdown10.0 KB
references/task-types.mdmarkdown9.9 KB
references/verification-guide.mdmarkdown10.2 KB
skill.ymlconfig6.0 KB
tests/test-triggers.jsonconfig6.3 KB

tests/test-triggers.json

View on GitHub
{
  "skill": "task-runner",
  "version": "2.0.0",
  "description": "Trigger test cases for the task-runner skill (v2.0.0 — two-mode design). Positive cases should activate the skill in the indicated mode; negative cases should NOT activate the skill.",

  "modes": {
    "INTAKE": "Triggered by user message containing task intent",
    "DISPATCHER": "Triggered by heartbeat poll or cron systemEvent"
  },

  "positive": [
    {
      "id": "PT-01",
      "mode": "INTAKE",
      "input": "Add these tasks for me: 1. Look up the weather in Paris 2. Create a summary file",
      "expected": "ACTIVATE",
      "reason": "Explicit 'add these tasks' phrase with numbered list"
    },
    {
      "id": "PT-02",
      "mode": "INTAKE",
      "input": "Do this for me: search for the latest AI news and summarize it in a markdown file",
      "expected": "ACTIVATE",
      "reason": "Explicit 'do this for me' trigger phrase"
    },
    {
      "id": "PT-03",
      "mode": "INTAKE",
      "input": "My task list for today:\n- Research competitor pricing\n- Write a pricing comparison report\n- Send the report to #research channel",
      "expected": "ACTIVATE",
      "reason": "Contains 'task list' phrase with bulleted multi-item list"
    },
    {
      "id": "PT-04",
      "mode": "INTAKE",
      "input": "Show me my tasks",
      "expected": "ACTIVATE",
      "reason": "'My tasks' trigger phrase — status check in INTAKE mode"
    },
    {
      "id": "PT-05",
      "mode": "INTAKE",
      "input": "Task status?",
      "expected": "ACTIVATE",
      "reason": "'Task status' trigger phrase — user wants current queue state"
    },
    {
      "id": "PT-06",
      "mode": "INTAKE",
      "input": "skip T-03",
      "expected": "ACTIVATE",
      "reason": "Task control command — skip T-NN pattern"
    },
    {
      "id": "PT-07",
      "mode": "INTAKE",
      "input": "retry T-02",
      "expected": "ACTIVATE",
      "reason": "Task control command — retry T-NN resets task to pending"
    },
    {
      "id": "PT-08",
      "mode": "INTAKE",
      "input": "cancel T-05",
      "expected": "ACTIVATE",
      "reason": "Task control command — cancel T-NN marks as skipped"
    },
    {
      "id": "PT-09",
      "mode": "INTAKE",
      "input": "I need you to pull the weekly sales numbers and create a summary report",
      "expected": "ACTIVATE",
      "reason": "'I need you to' trigger with clear multi-step task"
    },
    {
      "id": "PT-10",
      "mode": "INTAKE",
      "input": "task: verify the backup script ran successfully last night",
      "expected": "ACTIVATE",
      "reason": "Explicit 'task:' prefix trigger"
    },
    {
      "id": "PT-11",
      "mode": "INTAKE",
      "input": "Help me with these:\n1. Check disk usage on the server\n2. Clean up log files older than 30 days\n3. Send me a summary of what was deleted",
      "expected": "ACTIVATE",
      "reason": "'Help me with' trigger + numbered multi-task list"
    },
    {
      "id": "PT-12",
      "mode": "INTAKE",
      "input": "What's in my task queue?",
      "expected": "ACTIVATE",
      "reason": "'Task queue' status check trigger"
    },
    {
      "id": "PT-13",
      "mode": "INTAKE",
      "input": "Queue these tasks: fetch today's exchange rates, convert 100 units, and save the result",
      "expected": "ACTIVATE",
      "reason": "'Queue these tasks' trigger + three-item compound list"
    },
    {
      "id": "PT-14",
      "mode": "INTAKE",
      "input": "I need to generate a weekly report and send it to the team",
      "expected": "ACTIVATE",
      "reason": "'I need to' trigger with compound task (generate + send)"
    },
    {
      "id": "PT-15",
      "mode": "INTAKE",
      "input": "Can you do this: check if the API endpoint is responding, and if it is, run the integration tests",
      "expected": "ACTIVATE",
      "reason": "'Can you do this' trigger with conditional compound task"
    },
    {
      "id": "PT-16",
      "mode": "DISPATCHER",
      "input": "TASK_RUNNER_DISPATCH: check queue and run pending tasks",
      "expected": "ACTIVATE",
      "reason": "Explicit cron/heartbeat systemEvent for dispatcher mode"
    },
    {
      "id": "PT-17",
      "mode": "DISPATCHER",
      "input": "HEARTBEAT: task-runner dispatcher check",
      "expected": "ACTIVATE",
      "reason": "Heartbeat systemEvent for dispatcher mode"
    }
  ],

  "negative": [
    {
      "id": "NT-01",
      "input": "What is the capital of France?",
      "expected": "NO_ACTIVATE",
      "reason": "Single standalone question — answer directly, no task queue needed"
    },
    {
      "id": "NT-02",
      "input": "Remind me in 20 minutes to take a break",
      "expected": "NO_ACTIVATE",
      "reason": "Single scheduling request — handled by scheduling skill directly"
    },
    {
      "id": "NT-03",
      "input": "Search for best pizza recipes",
      "expected": "NO_ACTIVATE",
      "reason": "Single web search — answer directly without task overhead"
    },
    {
      "id": "NT-04",
      "input": "What time is it?",
      "expected": "NO_ACTIVATE",
      "reason": "Trivial single lookup — no task infrastructure needed"
    },
    {
      "id": "NT-05",
      "input": "Thanks, that looks great!",
      "expected": "NO_ACTIVATE",
      "reason": "Conversational acknowledgment — no action requested"
    },
    {
      "id": "NT-06",
      "input": "What's the weather like today?",
      "expected": "NO_ACTIVATE",
      "reason": "Single info lookup — answer directly, too trivial for task queue"
    },
    {
      "id": "NT-07",
      "input": "How do I use the message tool?",
      "expected": "NO_ACTIVATE",
      "reason": "Single documentation question — answer directly"
    }
  ],

  "notes": [
    "Trigger accuracy threshold: ≥90% (positive + negative combined)",
    "All 17 positive cases should activate the skill in the indicated mode",
    "All 7 negative cases should NOT activate the skill",
    "INTAKE mode: prefer activating when user intent is to add/manage tasks",
    "DISPATCHER mode: only activated by systemEvents, never by direct user messages",
    "Task control commands (skip/retry/cancel T-NN) always trigger INTAKE",
    "Status checks ('show tasks', 'task status') always trigger INTAKE",
    "When ambiguous between INTAKE and answering directly: if message has ≥2 distinct action items, activate INTAKE"
  ]
}
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 322kUpdated Claude CodeCodexVerifiedInstall