prime

Scan repository structure and load priority docs to generate a context summary.

156|37|Updated Apr 12, 2026
One-click install
npx skills add https://github.com/zeuikli/claude-code-workspace --skill prime-zeuikli
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: prime
Source: https://github.com/zeuikli/claude-code-workspace/tree/main/.claude/skills/prime
Command: npx skills add https://github.com/zeuikli/claude-code-workspace --skill prime-zeuikli

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill solves the problem of starting a new or unfamiliar codebase without spending lots of time manually figuring out its structure and purpose.

Core Features & Use Cases

  • Repository structure scanning: Quickly inspects tracked files and produces a directory/file overview to orient you.
  • Priority documentation loading: Reads README.md, then CLAUDE.md, then docs/INDEX.md (when present) to establish goals and conventions.
  • Context summary generation: Produces a compact ≤300-word project context including purpose, stack, key modules, and suggested next steps; also adds workspace-specific inventory for this repo (rules, skills, agents, and a healthcheck hint).

Quick Start

Use the prime skill to scan the current repo and generate a concise context summary for your next development task.

Frequently Asked Questions about prime

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

FAQPage Schema
How do I generate a context summary for an unfamiliar codebase?

To generate a codebase context summary, scan the repository's tracked file structure using git ls-files and load prioritized documentation like README.md to orient yourself quickly. This produces a compact overview of the project's purpose, stack, and modules.

What is the best way to bootstrap a new session after a context clear?

Bootstrapping a new session requires scanning tracked files and loading priority docs such as CLAUDE.md and docs/INDEX.md. This cold-start approach rapidly rebuilds your working context, providing purpose, stack details, and next-step guidance.

Does context loading read all files in the repository?

Context loading does not read all files; it enforces strict reading limits of no more than three full files and bounds the summary length. It scans the file structure via git while excluding .git and node_modules directories.

How do I scan a repository structure without reading every file?

You can scan a repository structure by listing tracked files with git ls-files and excluding directories like node_modules. This method provides a directory and file overview to orient you without hitting filesystem reading limits.

What documentation is prioritized during a repository context scan?

A repository context scan prioritizes loading README.md, then CLAUDE.md, and finally docs/INDEX.md if present. This targeted documentation loading establishes project goals and conventions efficiently for new sessions.

When do I need to run a codebase scan for project orientation?

You need a codebase scan for project orientation during new sessions, when tackling unfamiliar repos, after clear or compact operations, or in hands-off handover scenarios where fast orientation is required.