deepwork_jobs

Design, validate, and implement multi-step DeepWork AI workflows from job.yml specifications.

49|3|Updated Jan 11, 2026
One-click install
npx skills add https://github.com/Unsupervisedcom/deepwork --skill deepwork-jobs-unsupervisedcom
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: deepwork_jobs
Source: https://github.com/Unsupervisedcom/deepwork/tree/main/.claude/skills/deepwork_jobs
Command: npx skills add https://github.com/Unsupervisedcom/deepwork --skill deepwork-jobs-unsupervisedcom

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps product teams and software engineers design, validate, and implement multi-step AI workflows for DeepWork jobs, ensuring consistency from definition to deployment and enabling learning from execution to improve future runs.

Core Features & Use Cases

  • Define new multi-step workflows using the new_job workflow: define, review_job_spec, and implement.
  • Review and validate job specifications with a dedicated sub-agent to ensure quality.
  • Run standalone learn to capture learnings from conversations and improve future instructions.
  • Operate via slash commands like /deepwork_jobs.define to start workflows.
  • Context: job specifications and learnings are stored in files such as the .deepwork/jobs/deepwork_jobs/job.yml and AGENTS.md.

Quick Start

Start a new workflow by invoking the skill to define a DeepWork job. The system will guide you through structured questions to specify inputs, outputs, and steps, then generate the necessary files and commands to implement the job.

Frequently Asked Questions about deepwork_jobs

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

FAQPage Schema
How do I define and automate multi-step AI workflows?

To define and automate multi-step AI workflows, you invoke the skill to start a new job, answer structured questions for inputs, outputs, and steps, then generate the necessary job.yml specifications and slash commands for implementation.

What is the process for validating job specifications before implementation?

Validating job specifications involves using a dedicated sub-agent to review the structured job.yml file, ensuring quality and consistency from definition through deployment before you implement the workflow steps.

How do I capture learnings from past AI workflow runs to improve future jobs?

You capture learnings from past AI workflow runs by using the standalone learn command, which stores insights from conversations into an AGENTS.md file to improve future instructions and job execution.

Can I use slash commands to trigger AI workflow automation?

Yes, you can use slash commands like deepwork_jobs.define to trigger AI workflow automation, guiding you through specifying inputs and outputs to generate the required implementation files.

Where are AI workflow job specifications and learnings stored?

AI workflow job specifications and learnings are stored in structured files, specifically the .deepwork/jobs/deepwork_jobs/job.yml for specifications and AGENTS.md for capturing execution learnings.