# Chronicling America [Health: Active]

**Category:** 🔎 Search & Data Extraction  
**Repository:** https://github.com/pipeworx-io/mcp-chronicling-america  
**GitHub Stars:** 0  
**npm Downloads (last month):** 1454  
**Views:** 0  
**Installs:** 0  
**Upvotes:** 0  
**Directory Page:** https://allmcps.com/mcp/chronicling-america

## Description
Chronicling America MCP — full-text search of ~150 years of digitized U.S.

## Tools
Capabilities this server exposes over MCP:

- **search_newspapers** — Full-text search of digitized historical U.S. newspapers (~1690s–present) from the Library of Congress "Chronicling America" collection. Find primary-source newspaper coverage of any person, event, place, or topic in history — returns the total match count, then for each page the newspaper, publication date, city/state and a link to the digitized page (plus an OCR text excerpt of the matching page for covered states such as Montana). Filter by date range and U.S. state; every response carries data_as_of. Example: search_newspapers({ query: "influenza epidemic", date_from: "1918", date_to: "1919", state: "oklahoma" }).

## Claude Desktop Quick Installation
Install path detected from listing signals. Uses `npx` (confidence: high):

```json
"mcpServers": {
  "chronicling-america": {
    "command": "npx",
    "args": ["-y","@pipeworx/mcp-chronicling-america"]
  }
}
```

## Documentation

## What Chronicling America does

Chronicling America provides MCP access to full-text search over digitized U.S. newspapers held in the Library of Congress's Chronicling America collection. The search covers newspapers spanning roughly the 1690s to the present and is suited to historical research involving people, events, places, and topics.

Results identify the newspaper, publication date, city and state, and the corresponding Library of Congress page viewer link. Matching pages can also include an excerpt from the page's OCR text. Responses include the total number of matches and a `data_as_of` value so a caller can identify the data timestamp.

## How it works

The primary MCP tool is `search_newspapers`. It accepts a required text query and optional `state`, `date_from`, `date_to`, and `limit` parameters. State filters accept a state name or postal abbreviation; date filters can use a year or an exact date. Results are ranked by relevance.

Requests are handled through the Pipeworx gateway. Depending on the query and available coverage, the gateway uses its own search coverage or sends the request to the Library of Congress search service. Responses identify the serving path with `served_from`. Pipeworx search responses also include coverage details such as years, pages searched, and batches. National searches have an approximately seven-second bound, and partial coverage is identified in the response when applicable.

The gateway also supports `ask_pipeworx`, which lets an agent submit a plain-language question and have the gateway select and populate a tool call. Pack endpoints expose shared Pipeworx meta-tools alongside the pack tool; the exact tool list is returned during MCP initialization.

## Setup and configuration

A hosted MCP connection can be added with this endpoint:

```json
{
  "mcpServers": {
    "chronicling-america": {
      "url": "https://gateway.pipeworx.io/chronicling-america/mcp"
    }
  }
}
```

The hosted connection is keyless. It can also be called directly over HTTP at `https://gateway.pipeworx.io/v1/tools/search_newspapers` with a JSON POST body.

For local stdio operation, run the npm bridge with `npx`:

```json
{
  "mcpServers": {
    "chronicling-america": {
      "command": "npx",
      "args": ["-y", "@pipeworx/mcp-chronicling-america"]
    }
  }
}
```

The bridge forwards calls to the hosted gateway and therefore needs network access. Anonymous use is capped at 50 calls per day. Adding `PIPEWORX_API_KEY` raises the allowance to 200 calls per day; the key is supplied as a Bearer token. Set `PIPEWORX_META_TOOLS=1` if the local bridge should list the shared gateway tools as well as the newspaper search tool.

## Tools and capabilities

- `search_newspapers`: Search OCR text with optional state, date-range, and result-limit filters.
- Return total match counts before individual page results.
- Provide newspaper, date, city, state, and Library of Congress page-link metadata.
- Include matching page OCR excerpts where available.
- Report `data_as_of`, serving path, and coverage information.

## Limitations and notes

Coverage is being added state by state. Montana is covered by the Pipeworx search path, while other collection partitions may be routed to the Library of Congress search service or may contribute only partial results. A failure from that upstream search is reported as `upstream_down`, rather than being presented as an empty result set.

OCR excerpts are raw transcriptions and can be noisy, especially for nineteenth- and early-twentieth-century newspaper type. Search results link to the Library of Congress page viewer for source inspection. The integration is an independent, unofficial connection and is not affiliated with, endorsed by, or published by the Library of Congress.

_Full upstream README: https://allmcps.com/mcp/chronicling-america/readme_