expert-manager

Create, convert, validate, and register Sztubuddy expert packages for agents and teams.

68|25|Updated Jul 24, 2026
One-click install
npx skills add https://github.com/rojim666/SztuCode --skill expert-manager-rojim666
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: expert-manager
Source: https://github.com/rojim666/SztuCode/tree/main/py-runtime/src/sztu_code/core/prompts/workbuddy/skills/expert-manager
Command: npx skills add https://github.com/rojim666/SztuCode --skill expert-manager-rojim666

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Building a publishable AI expert package involves many strict conventions—plugin.json fields, Agent MD structure, team collaboration rules, avatar generation, and marketplace registration—that are easy to get wrong. This Skill manages the full lifecycle of expert packages so they pass validation and can be submitted for review. ## Core Features & Use Cases - Expert Creation & Conversion: Create Agent-type or Team-type experts interactively, or convert existing materials (docs, prompts, SOPs) into a compliant expert package. - Modification & Compliance: Edit existing experts (display info, tags, prompts, members) while protecting immutable identity fields, then re-validate and re-register. - Automated Pipeline: Initialize directories, generate avatars via ImageGen, validate with scripts, register into marketplace.json, and package for sharing. - Use Case: A user pastes an internal workflow document and asks to turn it into a three-member analysis team; the Skill infers expertType and categoryId, generates all files and avatars, then validates and registers the team. ## Quick Start Convert this document into a Sztubuddy expert package and register it to my expert marketplace.

Frequently Asked Questions about expert-manager

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

FAQPage Schema
How do I create an expert package from an existing document?▼

Provide the file path or paste the content, and the Skill extracts role definitions, capabilities, SOPs, and output rules, then infers expertType and categoryId for your confirmation. It then initializes the directory, generates all files and avatars, validates, and registers the expert.

What is the difference between Agent-type and Team-type experts?▼

Agent-type is a single AI expert with one MD file, while Team-type is a multi-role collaboration team with a lead agent and member agents. Team-type requires a teamInfo field, an SOP workflow in the lead's MD, and a settings.json file.

Can I rename an existing expert or change its directory name?▼

No. The name field, agentName, directory name, and MD filenames are unique identifiers and cannot be modified in place. Renaming requires creating a new expert; only display fields like displayName, tags, and descriptions are editable.

Where are generated experts stored?▼

Experts must be generated under $Sztubuddy_CONFIG_DIR/plugins/marketplaces/my-experts/plugins, defaulting to ~/.Sztubuddy/plugins/marketplaces/my-experts/plugins. Experts created elsewhere cannot be detected or used.

Why does expert registration fail after validation?▼

Registration re-checks that key fields contain no [TODO] placeholders and that validation passed. If any placeholder remains or required fields are malformed, the register script refuses to write to marketplace.json.