getting-started

Orients Claude Code sessions by listing available skills, agents, and workflow rules.

Updated Aug 13, 2026
One-click install
npx skills add https://github.com/Evolutionary-Leadership/harness --skill getting-started-evolutionary-leadership
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: getting-started
Source: https://github.com/Evolutionary-Leadership/harness/tree/main/.claude/setup/railway/.claude/skills/getting-started
Command: npx skills add https://github.com/Evolutionary-Leadership/harness --skill getting-started-evolutionary-leadership

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? New Claude Code sessions start without knowledge of the project's available skills, agents, and workflow conventions, leading to improvised approaches that bypass established procedures. This Skill runs at session startup to teach the AI what tools exist and how to use them correctly. ## Core Features & Use Cases - Skill and Agent Discovery: Runs list-skills.sh to enumerate every available skill and agent with descriptions, so the AI knows what procedures exist before acting. - Session Flavor Classification: Distinguishes chat, brainstorm, and feature sessions, routing each to the correct workflow skill (/chat, /brainstorm, /feature). - Response Conventions: Enforces a standardized closing block (Good to know / Act later / Act next) and a stand-down block for work that cannot be completed, keeping multi-session coordination consistent. - Use Case: A developer opens a fresh Claude Code session in a harness-managed repository; this Skill ensures the AI immediately discovers /feature, /code-review, and the documentation manifest instead of improvising. ## Quick Start Ask the AI to run the getting-started orientation so it lists all available skills and agents before starting work on your request.

Frequently Asked Questions about getting-started

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

FAQPage Schema
How do I see what skills are available in a Claude Code project?

Run bash .claude/scripts/list-skills.sh in the session. It lists every skill in .claude/skills/ and every agent in .claude/agents/ along with its description, so you know which procedures exist before starting work.

What is the difference between chat, brainstorm, and feature sessions?

Chat sessions write nothing, brainstorm sessions write only to the issue tracker as idea issues, and feature sessions modify the repository through five gated phases. The user states the flavor explicitly at session start, and each routes to its corresponding skill.

When should Claude Code use a skill instead of answering directly?

Whenever a skill or agent exists for the requested task, it must be used rather than improvised. The orientation rules treat skills as mandatory, and even a small chance that a skill applies requires checking the list first.

What is the stand-down block in Claude Code sessions?

It is a fixed HARNESS BLOCKED format emitted when a session cannot finish requested work due to missing authority, missing capability, or a needed decision. It records what was done, what remains, and the smallest action that would unblock completion.

Does the getting-started skill modify repository files?

No. It only reads the skill catalog via list-skills.sh and teaches behavioral conventions such as closing blocks and question numbering. File changes happen later through workflow skills like /feature or /brainstorm.