hug

Interview users and aggregate project metadata to create per-user profiles.

4|1|Updated Apr 11, 2026
One-click install
npx skills add https://github.com/nicolasguelfi/gensem --skill hug
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: hug
Source: https://github.com/nicolasguelfi/gensem/tree/main/gse-one/plugin/skills/hug
Command: npx skills add https://github.com/nicolasguelfi/gensem --skill hug

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Establish or refresh the per-user profile to enable contextual interactions when user context is unknown or missing, triggered by /gse:hug.

Core Features & Use Cases

  • Interactive interview to build or update a per-user profile based on project signals (.gse/profile.yaml, .gse/config.yaml, Git history, and package manifests).
  • Supports workflow flags like --update, --show, and --team to manage single or team contexts.
  • Persists the profile under .gse/profile.yaml and aligns with project state automatically.

Quick Start

Run hug to start or update the per-user profile for the current project

Frequently Asked Questions about hug

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

FAQPage Schema
How do I create a user profile for my project using Git history and package manifests?

Create a user profile by running hug, which triggers an interactive interview that aggregates signals from Git logs and package manifests like pyproject.toml or package.json, persisting the resulting context to .gse/profile.yaml.

What is the best way to update an existing per-user profile when project context changes?

The best way to update an existing per-user profile is to run hug with the --update flag, which refreshes project signals and aligns the stored context in .gse/profile.yaml with the current project state automatically.

Can I manage team contexts and view multiple user profiles in a single project?

Yes, you can manage team contexts by running hug with the --team flag to handle multi-user environments, and use the --show flag to display the currently stored profile configuration.

Does the profile management workflow require specific configuration files to infer language and domain context?

The profile management workflow reads .gse/config.yaml, Git logs, and domain manifests such as pyproject.toml, package.json, or Cargo.toml to infer language, domain, and team context without requiring additional manual setup.

When should I not use an interactive interview to establish project configuration?

You should avoid the interactive interview if the .gse/profile.yaml already exists and no project metadata has changed, because the profile aligns automatically with the project state and requires updates only when context is missing or stale.