An open API service indexing awesome lists of open source software.

https://github.com/twitu/indiana-codes

Read a codebase like a book — chapter-by-chapter git/PR/issue archaeology with interactive Q&A. Adds the /dig slash command for Claude Code.
https://github.com/twitu/indiana-codes

ai-agents claude-code claude-code-plugin claude-code-skill codebase documentation git-archaeology onboarding

Last synced: 5 days ago
JSON representation

Read a codebase like a book — chapter-by-chapter git/PR/issue archaeology with interactive Q&A. Adds the /dig slash command for Claude Code.

Awesome Lists containing this project

README

          

# Code Archaeology

> *Read the codebase like a book. Chapter by chapter. With a guide who answers back.*

A skill for AI coding agents (Claude Code, OpenCode, Cursor, …) that turns a repo's git history + GitHub issues/PRs/discussions into a chapter-based, interactive learning experience. You run `/dig`, the agent does the digging, and you come out the other side understanding the codebase like someone who's been on the team a year.

## You just joined the team

You open the repo. **2,000 commits.** 200+ tickets. 14 contributors. Half of them don't work here anymore. The README was last updated in 2023.

What do you do?

![wing-it](./wing-it.jpg)

You read the code. You find a decision that makes no sense. You want to ask your teammates but you've been here 3 days and they seem busy and you don't want to be *that* person.

What do you do?

![hack right?](./hack-right.jpg)

You be Indiana Jones \*cracks whip. You **dig**. You study the history, chapter by chapter, until you understand *why* things are the way they are — and you can challenge them with evidence.

![tame-the-code](./tame-the-code.jpg)

## What it does

1. **Survey** — commits, contributors, ticket patterns, infra files, gaps in documentation, the suspicious silence around `legacy_v2/`.
2. **Identify chapters** — natural breakpoints in the project's life. Team transitions. Migrations. The Great Refactor of 2022. The incident that shall not be named.
3. **Deep-dive** — for each chapter, walk you through what happened, who did it, why, and what got abandoned along the way.
4. **Quiz you back** — *why not the simpler thing?* *what breaks if this assumption is wrong?* *was this eventually replaced?* You answer. You learn. The chapter writes itself into `PROJECT_HISTORY.md` and `chapters/-.md`.

The Q&A is the load-bearing phase. Steps 1–3 produce a draft; step 4 — the conversation — is where you actually learn the system.

## Install

### Claude Code (one-liner via marketplace)

```
/plugin marketplace add twitu/indiana-codes
/plugin install dig@indiana-codes
```

That's it. Restart and `/dig` will appear in your slash-command menu in any repo with git history.

### Manual install (OpenCode, Cursor, anything else)

Clone the repo first. The skill content is identical across agents — only the install path differs.

```bash
git clone https://github.com/twitu/indiana-codes ~/.local/share/indiana-codes
cd ~/.local/share/indiana-codes
```

#### OpenCode

```bash
mkdir -p ~/.config/opencode/skills ~/.config/opencode/commands
ln -sf ~/.local/share/indiana-codes/dig ~/.config/opencode/skills/dig
ln -sf ~/.local/share/indiana-codes/commands/dig.md ~/.config/opencode/commands/dig.md
```

#### Cursor

Cursor has no skills directory, so the command file points the agent at the cloned `SKILL.md` directly:

```bash
mkdir -p ~/.cursor/commands
cat > ~/.cursor/commands/dig.md <