Indexandria

Indexandria

**Crawl any web documentation and pull it straight into Claude Code's context.**

Indexandria is a [Claude Code](https://code.claude.com) plugin that fetches documentation from any URL, converts it to clean markdown, and returns it directly into the conversation. Claude then uses this content as context while writing code — no database, no background server, no disk writes.

*Named after the Great Library of Alexandria.*

Install

/plugin marketplace add alicankiraz1/indexandria
/plugin install indexandria@indexandria

Requires Python 3.10+ and [uv](https://docs.astral.sh/uv/).

Quick Start

Tell Claude to fetch some docs:

> Crawl https://fastapi.tiangolo.com/tutorial/ and then build me an API with validation

For larger documentation sites, use deep indexing:

> Index the full React reference docs, then help me build a custom hook

Two Workflows

Quick Mode — Small, Focused Crawls

`crawl_docs` fetches 1-15 pages and returns content directly into context. Best for single API lookups and short tutorials.

"Crawl the useEffect docs"
       │
       ▼
 ┌───────────┐     ┌────────────┐     ┌───────────┐
 │  Claude    │────▶│ crawl_docs │────▶│  Crawler   │
 │  Code      │     │ (MCP tool) │     │ httpx+BS4  │
 └───────────┘     └────────────┘     └───────────┘
       │                                     │
       │          markdown content           │
       │◀────────────────────────────────────┘
       ▼
 Claude writes code using the docs

Deep Mode — Large Documentation Sites

`index_docs` crawls up to 100 pages into memory and returns a compact table of contents. Claude then uses `search_indexed` and `get_indexed_page` to selectively read only what's needed.

"Index the FastAPI tutorial, then build OAuth2 auth"
       │
       ▼
 ┌───────────┐     ┌────────────┐     ┌───────────┐
 │  Claude    │────▶│ index_docs │────▶│  Crawler   │
 │  Code      │     │            │     │ (100 pages)│
 └────────