execute

Execute planned task DAGs with isolated subagents and checkpoint recovery.

19|1|Updated Nov 25, 2025
One-click install
npx skills add https://github.com/Dowwie/tasker --skill execute-dowwie
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: execute
Source: https://github.com/Dowwie/tasker/tree/main/skills/execute
Command: npx skills add https://github.com/Dowwie/tasker --skill execute-dowwie

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Orchestrates end-to-end execution of a planned task DAG by spawning isolated subagents that run tasks deterministically, capture results, and surface artifacts for verification.

Core Features & Use Cases

  • Isolated task execution: spawn per-task subagents to ensure isolation and reproducibility.
  • Parallel execution with safety: supports up to 3 concurrent executors and built-in crash recovery.
  • Artifact management: writes bundles, task results, and evaluation reports under the .tasker directory for traceability.
  • Recovery and resilience: supports plan-to-execute workflow with checkpointing, halt/resume, and audit-ready state tracking.

Quick Start

Start by running /tasker:plan to generate a DAG of tasks, then run /tasker:execute to begin execution.

Frequently Asked Questions about execute

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I execute a task DAG with isolated subagents?

Task DAG execution supports up to three concurrent subagents. This parallelism limit ensures deterministic task execution while maintaining crash recovery through checkpoints for halt and resume operations.

How do I start an end-to-end spec-to-execute workflow?

Completed task execution writes bundles and evaluation reports to .tasker/bundles and .tasker/reports. These artifacts provide traceability and surface results for verification after the subagents finish running.

Does task execution support crash recovery and checkpoints?

Yes, task execution enforces crash recovery through checkpoints. It reads from .tasker/state.json for audit-ready state tracking, allowing you to halt and resume the workflow without losing progress on isolated tasks.

What do I need before running isolated task execution?

You need a completed plan with defined task definitions within a CLI-driven project. The execution process relies on this pre-existing plan and requires the .tasker directory structure to manage state, bundles, and reports.