automate-me

Drafts a personal mode skill by mining agent transcripts and user preferences.

Updated Sep 2, 2026
One-click install
npx skills add https://github.com/jnyross/pstack-muse --skill automate-me-jnyross
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: automate-me
Source: https://github.com/jnyross/pstack-muse/tree/main/skills/automate-me
Command: npx skills add https://github.com/jnyross/pstack-muse --skill automate-me-jnyross

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Capturing your working conventions into a reusable skill that agents follow is tedious and error-prone; this Skill automates the discovery, drafting, and refinement of a personal "-mode" skill so agents consistently work the way you do. ## Core Features & Use Cases - Transcript Mining: Runs parallel subagents over recent workspace agent transcripts to detect recurring patterns in response style, delegation habits, verification posture, and process conventions. - Structured Preference Elicitation: Uses structured multi-choice questions to capture intent that mining misses, then clusters findings into sections like autonomy, subagents, and review practices. - Guided Authoring and Iteration: Orchestrates the create-skill and unslop skills to draft, refine, and land the mode skill via a worktree, commit, and PR. - Use Case: After weeks of correcting agents on tone and workflow, ask to "automate me" and receive a drafted jay-mode skill grounded in evidence from your actual conversation history. ## Quick Start Ask the agent to automate me by turning my working style and preferences into a personal mode skill.

Frequently Asked Questions about automate-me

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

FAQPage Schema
How do I turn my working style into a skill for AI agents?

Invoke the automate-me flow, which mines your recent agent transcripts for recurring preferences, asks structured follow-up questions, and drafts a personal -mode skill. The draft is refined with prose guidelines and landed via a pull request for your review.

How does automate-me find my preferences from past conversations?

It runs parallel subagents across slices of your workspace's agent transcripts, looking for signals like response style, delegation habits, and verification posture. Patterns appearing in two or more slices are treated as high-confidence; one-off signals are dropped.

Can I update an existing mode skill instead of creating a new one?

Yes. If a matching -mode skill already exists, the flow defaults to update mode: it mines only history since the skill's last edit, asks what has changed, and edits the file in place while preserving sections you have not contradicted.

When should I use create-skill instead of automate-me?

Use create-skill alone when you want a task-specific skill or a single narrow workflow, such as how you write commit messages. Automate-me is for capturing broad working conventions into a personal mode skill and requires transcript mining.

Does automate-me read transcripts from other projects?

No. It only reads the active workspace's agent-transcripts directory named in the system prompt. It explicitly avoids globbing across other project directories to prevent crossing workspace boundaries and reading unrelated private chats.