init-hero

Investigates a repository and generates a HERO.md configuration file for all hero skills.

5|1|Updated Mar 7, 2026
One-click install
npx skills add https://github.com/ai-hero/hero-skills --skill init-hero-ai-hero
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: init-hero
Source: https://github.com/ai-hero/hero-skills/tree/main/skills/init-hero
Command: npx skills add https://github.com/ai-hero/hero-skills --skill init-hero-ai-hero

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Every development team configures its workflow differently: different PM tools, CI platforms, branch conventions, and code quality commands. Without a shared configuration, each automation skill would need to rediscover these settings on every run. This Skill investigates a repository, auto-detects the stack and conventions, confirms findings with evidence-based questions, and writes a single HERO.md file that all other hero skills read. ## Core Features & Use Cases - Deep Repository Investigation: Detects the tech stack, frameworks, package managers, task runners, CI/CD workflows, deployment targets, code review bots, PM tools, and coding conventions by scanning dependency files, git history, workflows, and source code. - Smart Confirmation Questions: Synthesizes findings into grouped, evidence-based questions instead of generic questionnaires, asking why only for unusual or non-default choices. - HERO.md and AGENTS.md Setup: Writes the HERO.md config consumed by skills like push-pr, review-pr, and ship-pr, and establishes AGENTS.md as the agent instructions file with CLAUDE.md as a symlink. - Recalibrate Mode: Re-investigates and refreshes an existing HERO.md when the configuration has gone stale. - Use Case: After cloning a new client repository with no HERO.md, run the init flow once so every subsequent hero skill knows the branch template, lint commands, CI platform, and review bot without re-asking. ## Quick Start Ask the agent to initialize Hero for this repository by investigating the stack and creating a HERO.md configuration file.

Frequently Asked Questions about init-hero

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

FAQPage Schema
How do I set up HERO.md for a new repository?

Run the init-hero skill on a repo with no HERO.md. It investigates the stack, CI, PM tools, and conventions, confirms findings with targeted questions, then writes HERO.md at the repo root for all other hero skills to use.

How do I update HERO.md when my project configuration changes?

Run init-hero with the recalibrate argument. It re-investigates the repository, compares findings against the existing HERO.md, and refreshes stale sections. The deprecated --update spelling is still accepted with a rename notice.

What does the init investigation detect in a repository?

It detects the tech stack and frameworks, package managers and lock files, task runners, lint and test commands, CI/CD workflows, deployment targets, code review bots like CodeRabbit or Greptile, PM tools from ticket prefixes, and coding conventions from source samples.

Does init-hero work with both CLAUDE.md and AGENTS.md?

Yes. It establishes AGENTS.md as the real instructions file and makes CLAUDE.md a symlink to it, handling states like CLAUDE-only, both-files, or wrong symlink targets. On Windows without symlink support it falls back to a pointer file.

What happens if my repo already has a HERO.md file?

Without the recalibrate argument, the skill shows the current configuration and asks whether you want to update it rather than overwriting. With recalibrate, it reads the existing file and refreshes it against new investigation findings.