Phase 1: Load Project Context

Phase 1:

Load Project Context

What Is This

The "Phase 1: Load Project Context" skill is the initial step in the onboarding process for new contributors or AI agents joining a project on the Happycapy Skills platform. This skill is essential for establishing a strong foundational understanding of a project's structure, standards, and current state. It accomplishes this by programmatically reading the CLAUDE.md file, which provides an authoritative project overview and codified standards. If a specific role or area is specified-such as programmer, designer, narrative, QA, or production-the skill also reads the corresponding agent definition in .claude/agents/ to tailor the onboarding experience.

Why Use It

Modern software projects are increasingly complex and collaborative, often involving distributed teams and AI-powered tools. Rapid onboarding-especially when powered by accurate, context-rich automation-minimizes ramp-up time and prevents costly misunderstandings. By consolidating project essentials in a repeatable, systematic way, this skill ensures new team members or AI agents can contribute effectively from day one. Loading the project context early prevents knowledge silos and establishes a uniform baseline for all contributors.

How to Use It

Invocation:
This skill is user-invocable and expects an argument specifying the relevant role or area (e.g., programmer, designer, QA, production). The skill is compatible with tools such as Read, Glob, Grep, and Write, enabling it to fetch and parse the required documentation files.

Process:

1. Read Project Overview:

   • The skill starts by reading the CLAUDE.md file located at the project root. This file contains a high-level summary of the project, its goals, architectural standards, and development conventions.

2. Role-Specific Context:

   • If a specific role is provided, the skill then reads the relevant agent definition file in .claude/agents/ (e.g., .claude/agents/programmer.md). This file outlines the expectations and responsibilities for that role, including any specialized standards or workflows.

3. Context Aggregation:

   • The extracted information is consolidated to form a contextual onboarding document. This includes the project summary, role responsibilities, architectural overview, and references to key directories and files.

Example Usage:

happycapy onboard programmer

This command generates a contextual onboarding summary for a new programmer, drawing from both the main project documentation and the role-specific agent definition.

When to Use It

• Onboarding New Contributors:
  Whenever a new human contributor or AI agent joins the project, use this skill as the very first onboarding step.
• Role Transition:
  If an existing team member switches roles or takes on new responsibilities, rerun this skill with the new role argument to refresh their context.
• AI Agent Initialization:
  When spinning up an AI agent to automate or assist with specific project tasks, ensure this skill is executed to provide accurate context.

Important Notes

• Strict Output Rules:
  The generated onboarding documentation must follow clear formatting and style rules-English only, 700-800 words, use H2 headers, and maintain a professional technical writing tone.
• No Em-Dashes or Emojis:
  To ensure clarity and parsing consistency, the output must avoid em-dashes and emojis.
• Context Completeness:
  This skill does not perform scanning of source code, design files, or test directories; those actions are handled in "Phase 2: Scan Relevant Area." Phase 1 is strictly for loading high-level context and standards.
• Role-Specific Customization:
  If no role or area is provided, only the general project context from CLAUDE.md is loaded. For maximum onboarding precision, always specify the relevant role or area.
• Security and Consistency:
  Only documented and version-controlled files are read, ensuring that the onboarding process is both reproducible and secure from accidental leaks of sensitive or outdated information.

Code Example

Below is a conceptual pseudocode snippet showing how this skill might be implemented:

def load_project_context(role=None):
    context = read_file('CLAUDE.md')
    if role:
        agent_file = f'.claude/agents/{role}.md'
        if file_exists(agent_file):
            agent_context = read_file(agent_file)
            context += '\n' + agent_context
    return context

Conclusion

The "Phase 1: Load Project Context" skill is a foundational component for efficient and reliable onboarding within the Happycapy Skills platform. By programmatically aggregating project overviews and role-specific standards, it ensures every new contributor or agent receives a concise, accurate, and actionable briefing. This reduces onboarding friction and lays the groundwork for more advanced, role-tailored onboarding and productivity workflows in subsequent phases. For best results, always begin the onboarding process by running this skill with the appropriate role or area argument.