Execute tracked tasks or plan files one phase at a time.
npx skills add https://github.com/mulev/solo-dev --skill execute-mulev
Please help me install this Agent Skill. Skill: execute Source: https://github.com/mulev/solo-dev/tree/main/skills/execute Command: npx skills add https://github.com/mulev/solo-dev --skill execute-mulev
Executes work from tracked tasks or plan files. Finds ready-to-work tasks via the configured issue tracker (priority and dependency aware), claims them, implements phases (write code, run tests, update plan), handles failures via bug rounds, and closes tasks on completion. Use when the user says "execute", "work on next task", "execute the plan", "start implementing", "run the plan", or passes a task ID or plan file path. Each invocation executes one phase — after completion, offers to continue with the next unblocked task. Also triggers on "resume", "what's next", or "next task". Accepts no argument (pick from ready tasks), a task ID, or a .md plan file path.
Instruct me to execute the next available phase from a plan or tracker task.
High-intent search queries and answers about installing and using this skill.
Task execution automation works by finding ready tracker tasks, claiming them, writing code, running tests, and closing them. You simply instruct the system to execute, and it handles one phase at a time.
Phase-by-phase plan execution implements work one phase at a time from a plan file or tracker task. After completing a phase, it prompts you to continue with the next unblocked task.
You can execute plan files without a tracker by passing a .md plan file path or a task ID. The system loads instructions and templates from plan files and references directly.
Priority and dependency awareness evaluates ready-to-work tasks from the configured tracker. It selects unblocked tasks based on priority, claims them, and processes them sequentially.
When tests fail during automated plan execution, it handles failures via bug rounds. It attempts to resolve the issues before continuing or closing the task.
You do not need a specific tracker integration to run the next available task. The system works with a configurable tracker, or you can pass a task ID or .md plan file path directly.
gioeAutomates backlog processing by dispatching prioritized tasks to /chain or /tusk endpoints until exhausted.
antinomyhqExecute structured task plans with automatic status tracking.
bordenetValidate issue comments against logs and source data before posting.
u9401066Automate single-task workflows through PLAN, EXECUTE, VERIFY, and COMPLETE phases.
