skills.rest
hieutrtr
hieutrtrCOMMUNITY
Explore all skills by this author →
GitHub

/project:skill

Generates Claude Code skill files (.claude/commands/*.md from detected domain evidence.

Updated Apr 3, 2026
Search Tags:#skill generation#Claude Code#domain detection#repository scanning#experience learning#evidence grounding#prompt automation
One-click install
npx skills add https://github.com/hieutrtr/claude-forge --skill project-skill
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: /project:skill
Source: https://github.com/hieutrtr/claude-forge/tree/main/.claude/commands
Command: npx skills add https://github.com/hieutrtr/claude-forge --skill project-skill

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It turns your existing repository into actionable Claude Code skills by scanning for real domain evidence (routes, models, tests, and vocabulary) and generating targeted command-style skill files you can run in Claude Code.

Core Features & Use Cases

  • Domain detection from code evidence: identifies the best-fit domain by scanning endpoint files, schema/model files, tests, and README noun phrases.
  • Experience-aware generation: uses prior /project:evolve outputs from .claude/forge-experience.md to avoid duplicates and to prioritize known missing skills or improve flagged ones.
  • Skill file synthesis with grounded evidence: generates .claude/commands/*.md skills whose templates include concrete file/method/route evidence rather than placeholders.
  • Safe iteration controls: supports --dry-run for previewing generated content and --domain NAME to force a domain when detection is uncertain.

Quick Start

Run /project:skill residential --dry-run to preview the domain-detected skill files that would be generated for a residential domain without writing anything.

Frequently Asked Questions about /project:skill

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

FAQPage Schema
How do I generate reusable Claude Code prompt commands from my existing repository?▼

You can generate Claude Code skills by scanning your repository for domain evidence like routes, models, and tests, then synthesizing reusable prompt command files automatically without manual authoring.

What is automatic domain detection for codebase prompt generation?▼

Domain detection identifies the best-fit domain by scanning endpoint files, schema models, tests, and README noun phrases, ensuring generated skill files contain concrete codebase evidence rather than placeholders.

How do I preview generated skill files without modifying my repository?▼

You can preview generated skill files by running the command with the --dry-run flag, allowing you to review the domain-detected prompt content before writing anything to your repository.

Can I force a specific domain when generating project skills if detection is uncertain?▼

Yes, you can force a specific domain by passing the --domain NAME option, overriding automatic domain detection when scanned repository evidence is uncertain or ambiguous.

How does experience learning improve prompt automation for existing projects?▼

Experience learning improves prompt automation by reading prior outputs from .claude/forge-experience.md to avoid duplicate skills, prioritize missing capabilities, and refine flagged command files.

Related Skills

Explore More Software Engineering Skills→
kapilvirenahujakapilvirenahuja
3

infer-domain-selection-from-code

Infers product domains from scan-index.json signals using canonical domain knowledge base.

Community
Intermediate
evolution-foundationevolution-foundation
520

create-command

Create Claude Code slash commands by generating .claude/commands/{name}.md files.

Official
Basic
grandameniumgrandamenium
88

tool-use-patterns

Coordinate Claude tool calls across sequential, parallel workflows with robust error handling.

Community
Advanced
LazyIsEfficientLazyIsEfficient
15

library-investigator

Probe Claude library files against RULESET.md and report violation counts.

Official
Advanced