Can PlainHub search across multiple repositories?

Yes. In All repos mode, you can search across all repositories accessible to the logged-in user. Filenames and content are searched simultaneously, and results are grouped by repository.

Search

Q: What is GitHub RAG in PlainHub?

Use all your GitHub repositories as a RAG data source. Just ask a question in natural language, and AI will search across repositories and answer with citations from relevant files (Coming Soon).

3 Search Modes

📄

Name

Search by filename

🔍

Content

Search within file contents

🌐

All repos

Search across all repositories

Cross-Repository Search (All repos)

Search across files scattered across GitHub, breaking through repository boundaries.

Scope: All repositories accessible to the logged-in user
Method: Simultaneous filename and content search (GitHub Search API + Trees API)
Results: Grouped by repository. Format: "X results / Y repositories"
Filtering: Filter by repository to show results from specific repos only

Use cases

Find information across documents scattered in multiple repositories
Search for specific keywords (API names, config values) across projects
Instantly answer "which repo was that file in?"

Toward a cross-repository knowledge server

Cross-repository search is the foundation for turning your GitHub repositories into an organizational knowledge server. By connecting with AI, it will support not just keyword search but natural language queries (RAG).

GitHub RAG — Cross-Repository Search × AI Coming Soon

Use all your GitHub repositories as a RAG (Retrieval-Augmented Generation) data source. Ask a question in natural language, and AI will search across repositories and answer with citations from relevant files.

"Where was the auth design doc again?"

→ AI finds related documents across 3 repositories:
  - repo-a/docs/AUTH_DESIGN.md — OAuth design policy
  - repo-b/SECURITY.md — Token management policy
  - repo-c/ADR/003-auth.md — Authentication ADR

PlainHub's cross-repository search already has the Retrieval foundation. By connecting it to the AI panel, a private RAG within your own GitHub becomes possible.

Service	Data source	Search method
ChatGPT / Claude	The entire web	General search
GitHub Copilot	Current repository	Code completion
PlainHub	All your GitHub repos	Cross-repo search × AI

Single-Repository RAG — Use GitHub as a database

Beyond cross-repository search, single-repository RAG is also a powerful use case. Use a GitHub repository as a database and let AI search and answer from its contents.

Example: Stock database

Store stock data and company profiles as Markdown/CSV in a GitHub repo, then query via natural language from PlainHub's AI panel.

stock-db/
├── companies/
│   ├── 7203_Toyota.md           ← Viewable directly in PlainHub
│   ├── 9984_SoftBank_Group.md   ← 1 file = 1 RAG chunk (optimal size)
│   └── ...
├── index.csv                    ← Search index
└── .github/workflows/           ← Auto-update via GitHub Actions

Example queries:
  "How are Toyota's recent financials?"
  "Which auto sector stocks have high dividend yields?"
  "Compare Sony and Toyota's P/E ratios"

Data sources: Wikipedia (business overview) + yfinance (financial data) + stock exchange (company master) — consolidated into one Markdown file per company on GitHub.

Benefits: No server needed, version-controlled, auto-updatable via GitHub Actions. Browse, edit, and query AI — all within PlainHub.