agentsclimarketplace

Onboarding

Skill RealDougEubanks/ClaudeMarketplace/skills/onboarding

A community-driven collection of custom skills for Claude Code, Anthropic's CLI tool for software engineering with Claude.

Install
npx -y skills add RealDougEubanks/ClaudeMarketplace --skill onboarding

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Generates a comprehensive developer onboarding guide (ONBOARDING.md) by reading the codebase: directory map, entry points, environment variables, key commands, and architecture overview.

SKILL.md

5.4 KB, as published. Nobody here has run it

Onboarding

Generate a comprehensive developer onboarding guide for an unfamiliar codebase and write it to docs/ONBOARDING.md.

Instructions

  1. Use Read on README.md, then attempt to read whichever of the following exist: package.json, pyproject.toml, go.mod, CLAUDE.md, docs/assumptions.md.

  2. Use Glob to map the top-level directory structure (*). For each top-level folder found, describe its purpose based on its name and contents (e.g., src/ → application source, scripts/ → automation scripts, docs/ → project documentation).

  3. Use Glob to find entry points matching: index.*, main.*, server.*, app.*, cmd/**/*. Read the primary entry point to understand how the application starts.

  4. Use Read on any Makefile, package.json (scripts section), or Taskfile.yml to identify key commands: install, run, test, build, lint.

  5. Use Glob to find .env.example or any env var documentation files. List all required environment variables with their descriptions.

    SECURITY: Read .env.example only — never read .env or any file containing real credential values, and never include actual secret values in the generated guide. Document variable names and descriptions only.

  6. Use Bash to count files by detected language type. Adapt the command to the language detected:

    • TypeScript/JavaScript: find . -name "*.ts" -o -name "*.js" | grep -v node_modules | wc -l
    • Python: find . -name "*.py" | grep -v __pycache__ | wc -l
    • Go: find . -name "*.go" | wc -l
  7. Generate architecture diagram.

    a. Use Read on the primary entry point file identified in step 3 to understand how the application initializes and what it connects to.

    b. Use Glob to find configuration files that reveal infrastructure: docker-compose.yml, *.env.example, infrastructure/**, terraform/**, serverless.yml, .github/workflows/**.

    c. Use Grep to identify major inter-module dependencies: import/require/use statements at the top of the entry point and primary route/handler files.

    d. Generate a Mermaid diagram using the most appropriate type:

    • For web apps with a request flow: sequenceDiagram showing Client → Load Balancer → App Server → Database/Cache/Queue
    • For microservices or modular apps: graph LR showing service dependencies
    • For simple single-module apps: graph TD showing function/module call flow

    e. Include the diagram in the generated docs/ONBOARDING.md under a "## Architecture" section, wrapped in a fenced code block with mermaid syntax tag.

    f. Add a note: "This diagram was auto-generated and may not capture all relationships. Review and update as the architecture evolves."

    Example output to include in ONBOARDING.md:

    ## Architecture
    
    ```mermaid
    sequenceDiagram
        Client->>+ALB: HTTPS Request
        ALB->>+AppServer: Forward
        AppServer->>+PostgreSQL: Query
        PostgreSQL-->>-AppServer: Result
        AppServer->>+Redis: Cache write
        AppServer-->>-ALB: Response
        ALB-->>-Client: HTTPS Response
    ```
    
    > This diagram was auto-generated and may not capture all relationships. Review and update as the architecture evolves.
    
  8. Use Glob to check for docs/agentRoster.md. If it exists, read it and include the agent roster section in the guide.

  9. Write the completed onboarding guide to docs/ONBOARDING.md using Write, following the output structure below.

Output Structure

Write the following document to docs/ONBOARDING.md:

# Developer Onboarding Guide — <Project Name>

## What This Project Does
<1-2 sentence summary from README>

## Tech Stack
| Layer | Technology |
|-------|-----------|
| Language | ... |
| Framework | ... |
| Database | ... |
| CI/CD | ... |

## Directory Map
| Path | Purpose |
|------|---------|
| `src/` | ... |

## Getting Started
```bash
# 1. Install dependencies
npm install

# 2. Configure environment
cp .env.example .env
# Edit .env: fill in ...

# 3. Run locally
npm run dev

Key Entry Points

  • src/index.ts:1 — Application bootstrap
  • src/routes/ — API route definitions

Environment Variables

VariableRequiredDescription
DATABASE_URLYesPostgreSQL connection string

Key Commands

CommandWhat it does
npm run devStart dev server with hot reload
npm testRun test suite

Architecture Overview

<brief description of the main data/request flow>

Agent Roster

<include only if docs/agentRoster.md exists>

How to Contribute

<derive from the target project's own conventions — see note below>


For the "How to Contribute" section: read the target project's `CONTRIBUTING.md`, `CLAUDE.md`, or `.github/PULL_REQUEST_TEMPLATE.md` and summarize *that project's* branching, review, and merge rules. Do not assume any particular branch-prefix or approval policy. If no convention docs exist, fall back to generic guidance: "Create a branch, open a PR, and request review — confirm the team's specific conventions with a maintainer."

After writing the file, confirm to the user that `docs/ONBOARDING.md` has been created and summarize the key sections included.

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.