init

Initialize the copydesk data directory and extract a voice register from writing samples.

81|4|Updated Apr 10, 2026
One-click install
npx skills add https://github.com/TimSimpsonJr/copydesk --skill init-timsimpsonjr
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: init
Source: https://github.com/TimSimpsonJr/copydesk/tree/main/skills/init
Command: npx skills add https://github.com/TimSimpsonJr/copydesk --skill init-timsimpsonjr

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Setting up a personalized writing assistant requires capturing your unique voice, which is hard to do manually. This Skill bootstraps the copydesk user-data directory and walks you through a two-pass extraction over your own writing samples to produce a reusable voice register. ## Core Features & Use Cases - Idempotent bootstrap: Creates the required data directories and copies template files without ever overwriting existing user data. - State-aware routing: Detects whether you have existing registers and routes you to add a new one or re-extract an existing one. - Two-pass voice extraction: Dispatches Sonnet agents over your writing samples and baseline Claude outputs to derive vocabulary, sentence structure, rhetorical techniques, and voice qualities, saved as a register file with trigger contexts. - Use Case: You want Claude to draft advocacy copy in your own voice. Run the init flow, provide 10-20 of your past pieces plus baseline samples, and receive a ready-to-use advocacy register that activates on matching writing contexts. ## Quick Start Ask Claude to run /copydesk:init to set up copydesk and create your first voice register from your writing samples.

Frequently Asked Questions about init

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

FAQPage Schema
How do I set up copydesk for the first time?

Invoke /copydesk:init. It creates the data directory at ~/.claude/data/copydesk/, copies missing template files, and walks you through extracting your first voice register from your writing samples.

How do I create a voice register from my own writing?

Provide 10-20 samples of your own writing plus 10 baseline Claude-generated samples on similar topics. The skill runs a two-pass extraction with Sonnet agents and saves the result as a register file with trigger contexts.

Can I add multiple voice registers for different writing styles?

Yes. Re-running /copydesk:init detects your existing registers and asks whether you want to add a new register or re-extract an existing one. Each register gets its own kebab-case filename and trigger list.

Will running init again overwrite my existing data?

No. The bootstrap phase only copies template files that are missing and never overwrites existing files. Existing registers are only replaced if you explicitly choose re-extraction for that register.

What if I don't have baseline writing samples ready?

The skill walks you through generating them: open a fresh Claude conversation with no special instructions, ask for 10 short pieces on topics similar to your samples, and save those outputs as the baseline corpus.