# CONTENTdm State Archives (pypi · contentdm-mcp)

Search US state archives' record images on CONTENTdm; read items, pages and transcripts.

- Trust score: 79/100 (medium)
- Change this week: +44
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-10-18

## Components

- pypi · `contentdm-mcp`: 79/100 (this document), [markdown](https://verifymcp.io/servers/ianderso-contentdm-mcp/contentdm-mcp.md), [page](https://verifymcp.io/servers/ianderso-contentdm-mcp/contentdm-mcp)

## Channel facts

- Registry: `pypi`
- Package: `contentdm-mcp`
- Version: `0.1.0`
- Transport: `stdio`

## Trust breakdown

How this component scores in each security and reliability category. Every signal is checked automatically from public evidence about the published package, including repeated runs of it in an isolated sandbox, and we only credit what we can confirm. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog)

Scored 2026-10-18.

- **Supply Chain Security**: 100/100
  - No malware found by supply-chain analysis.
  - No known CVEs affecting this package version or its production dependencies.
  - Runs hatchling.build at install time, a recognised build step with no custom scripting around it.
  - 1 of 33 dependencies flagged as unhealthy.
- **Provenance & Transparency**: 100/100
  - Source repository is publicly reachable at the declared URL.
  - Cryptographically verified build provenance (signed, bound to ianderso/contentdm-mcp).
  - Clear OSI-approved license (MIT).
  - Actively maintained (last published 11 days ago).
  - Publishes a security disclosure policy (SECURITY.md).
- **Schema Quality & AI Usability**: 60/100
  - AI-judged instruction clarity (good).
  - Context-footprint check failed: tool/resource definitions use about 1784 tokens (~223/item across 8 items; 8 tools + 0 resources), over budget; trim descriptions and params.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 0/100
  - Stability not yet verified: not enough scan history yet (needs a 30-day window).
- **Tool Coverage**: 100/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 100% of tool parameters carry a description.
- **Tool Safety**: 100/100
  - No prompt-injection markers were found in the server instructions, tool names or descriptions we captured.
  - We read all 8 captured tool definition(s), and no name or description among them implies an irreversible operation.
  - An AI judge read all 9 captured unit(s) of tool text and found none that tries to manipulate the model reading it.
- **Capabilities**: 100/100
  - Implements a current MCP spec version (2026-07-28).

**Unverified: 1 category.** A category scored 0 because we could not verify it: a data source with nothing on this package, evidence we could not reach, or a check we could not run. We only credit what we can confirm.

## Install

### How do I install the CONTENTdm State Archives MCP server?

CONTENTdm State Archives runs locally as a PyPI package, launched with uvx contentdm-mcp. Ready-made configuration for Claude, Cursor, VS Code, Codex and 5 more is on this page, copied from each client's own documentation.

### Claude

```bash
claude mcp add ianderso-contentdm-mcp -- uvx contentdm-mcp
```

### Cursor

```json
{
  "mcpServers": {
    "ianderso-contentdm-mcp": {
      "command": "uvx",
      "args": [
        "contentdm-mcp"
      ]
    }
  }
}
```

### VS Code

```json
{
  "servers": {
    "ianderso-contentdm-mcp": {
      "command": "uvx",
      "args": [
        "contentdm-mcp"
      ]
    }
  }
}
```

### Codex

```bash
codex mcp add ianderso-contentdm-mcp -- uvx contentdm-mcp
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "ianderso-contentdm-mcp": {
      "type": "local",
      "command": [
        "uvx",
        "contentdm-mcp"
      ],
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add ianderso-contentdm-mcp --command uvx --arg contentdm-mcp
```

### Hermes

```yaml
mcp_servers:
  ianderso-contentdm-mcp:
    command: "uvx"
    args: ["contentdm-mcp"]
```

### Netclaw

```json
{
  "McpServers": {
    "ianderso-contentdm-mcp": {
      "Transport": "stdio",
      "Command": "uvx",
      "Arguments": [
        "contentdm-mcp"
      ]
    }
  }
}
```

### Vellum

```bash
assistant mcp add ianderso-contentdm-mcp -t stdio -c uvx -a contentdm-mcp
```

### Other

```json
{
  "mcpServers": {
    "ianderso-contentdm-mcp": {
      "command": "uvx",
      "args": [
        "contentdm-mcp"
      ]
    }
  }
}
```

## Changelog

Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims.

### 2026-10-14 (score 79, +29)

- [security improvement] Injection markers: unverified → pass
- [security] First check of Judged manipulation: pass
- [security] Stability: Stability not yet verified: not enough scan history yet (needs a 30-day window).
- [functional improvement] Tool coverage: unverified → 100
- [functional improvement] MCP protocol: unverified → pass
- [functional] First check of Schema quality: fail
- [functional] First check of Destructive annotations: pass
- [functional] First check of Schema quality: fail
- [functional] First check of Schema quality: good
- [functional] First check of Tool coverage: 100

### 2026-10-12 (score 50, +15)

- [security improvement] Malware scan: unverified → pass

### 2026-10-08 (score 35)

First indexed and scored.

## MCP tools (8)

### `list_instances` (~137 tokens)

List the curated CONTENTdm sites, and any the operator added: who runs each, what it holds.

Makes no network call. Each `instance` value is what the other tools take.
Sites that have left CONTENTdm stay listed with where they went. Any other
CONTENTdm site works by its https://cdmNNNNN.contentdm.oclc.org address,
which every site has; other hosts are refused.

Input parameters:

- `include_unsupported` (boolean): Also list sites that have left CONTENTdm or are blocked.
- `state` (string): Two-letter state codes to keep, e.g. 'TN' or 'GA,AL'.

### `list_collections` (~126 tokens)

List a site's collections and the alias each is searched by.

Use the alias, not the short number some sites put in their web
addresses. A site that lists no collections has usually moved elsewhere.

Input parameters:

- `instance` (string, required): A key from list_instances, or an https://NAME.contentdm.oclc.org address.
- `name_contains` (string): Keep collections whose name has this.
- `refresh` (boolean): True asks the site again instead of using the cache, and replaces the cached copy. Searches are kept a day, items 7 days, collection lists 30.

### `get_collection` (~147 tokens)

Read one collection's fields: what search's `field` can name, and which hold text.

Field names are per collection: the same label has different nicks on
different collections, and `full_text` (a transcript or OCR) is marked
by type, whatever it is called.

Input parameters:

- `collection` (string, required): The collection's alias, from list_collections.
- `instance` (string, required): A key from list_instances, or an https://NAME.contentdm.oclc.org address.
- `refresh` (boolean): True asks the site again instead of using the cache, and replaces the cached copy. Searches are kept a day, items 7 days, collection lists 30.

### `search` (~379 tokens)

Search one CONTENTdm site, or several side by side, for items or pages.

\`answered`, `failed`, `timed_out` and `not_searched` say which sites the
answer covers: a negative is only as wide as `answered`. Matching covers
metadata and any transcript or OCR field, never the image itself, and
many record images have no text at all. A compound object's text is on
its pages: use page_hits=true, or get_pages with a query, to find the
page. Hits are leads: open the item and read the image.

Input parameters:

- `collection` (string): An alias, with one instance only.
- `count` (integer): Hits per site: up to 50 for one, 5 for several.
- `field` (string): Search one field by its nick (get_collection); default all.
- `instance`: One site (a key, or an https://NAME.contentdm.oclc.org address), a list of them, or omit for every supported curated site.
- `mode` (string): all: every word; any: any word; exact: the words as a phrase.
- `page` (integer): Page of results, 1-based.
- `page_hits` (boolean): True returns the matching pages inside compound objects instead of the objects: where a transcript or OCR match actually is.
- `query` (string, required): Words to find, e.g. 'alvin york'. Empty lists everything (one collection only). A trailing * matches word beginnings.
- `refresh` (boolean): True asks the site again instead of using the cache, and replaces the cached copy. Searches are kept a day, items 7 days, collection lists 30.
- `state` (string): With no instance, only sites in these states, e.g. 'TN,GA'.

### `get_item` (~281 tokens)

Read one item: its metadata, any transcript or OCR text, and how to cite it.

\`text` is a lead, not the record: read the image (get_image). A compound
object's own text is usually empty; its pages carry it (get_pages). For a
page, `page_of` names the object and page number. `cite_as` is the
institution's own citation where it gives one; `rights` its terms. An
address on an unlisted site's own domain is refused: use its
cdmNNNNN.contentdm.oclc.org form.

Input parameters:

- `collection` (string): The collection alias.
- `instance` (string): A key from list_instances, or an https://NAME.contentdm.oclc.org address.
- `pointer` (string): The item's number, e.g. '16539'.
- `refresh` (boolean): True asks the site again instead of using the cache, and replaces the cached copy. Searches are kept a day, items 7 days, collection lists 30.
- `text_chars` (integer): Transcript or OCR characters to return per field; 0 for none.
- `url` (string): Instead of the three above: the item's address, e.g. from DPLA's isShownAt (/digital/collection/... or an old /cdm/... form).

### `get_pages` (~238 tokens)

List a compound object's pages in order, each with its own number and address.

A volume, a file or a newspaper issue is one compound object of many
pages, and each page is an item with its own pointer. A PDF page can
itself hold several pages: get_image's pdf_page reaches them. `query`
marks which pages match, which is how to find a name in a long index.

Input parameters:

- `collection` (string): The collection alias.
- `count` (integer): Pages to list (1-500).
- `first` (integer): First page to list, 1-based.
- `instance` (string): A key from list_instances, or an https://NAME.contentdm.oclc.org address.
- `pointer` (string): The compound object's number.
- `query` (string): Mark the pages whose text or metadata match these words.
- `refresh` (boolean): True asks the site again instead of using the cache, and replaces the cached copy. Searches are kept a day, items 7 days, collection lists 30.
- `url` (string): Instead of the three above: the object's address.

### `get_image` (~273 tokens)

Download one page image through the site's IIIF service, so it can be read.

This is the step that turns a hit into evidence. The image is written as
a JPEG to `destination`, never returned inline: a new .jpg or .jpeg file,
not hidden, not under ~/Library, and inside the operator's download
folder when one is set. Audio and video cannot be rendered. Cite the page
by the `citation` returned, not by the file.

Input parameters:

- `collection` (string): The collection alias.
- `destination` (string, required): Absolute path of a new .jpg file, e.g. '/path/to/images/page-461.jpg'. The directory must exist and the file must not: nothing is overwritten.
- `instance` (string): A key from list_instances, or an https://NAME.contentdm.oclc.org address.
- `max_pixels` (integer): Longest side in pixels; 0 for the largest the site will render.
- `page`: For a compound object: which page, 1-based. Default 1.
- `pdf_page`: For a multi-page PDF item: which of its pages, 1-based.
- `pointer` (string): The item's or page's number.
- `url` (string): Instead of the three above: the item's address.

### `cache_status` (~26 tokens)

Report this session's requests, by site, and cache use. Makes no network call.

## Diagnostics

Captured diagnostic sections: Provenance, Install scripts, Dependencies. The full working is on the page: https://verifymcp.io/servers/ianderso-contentdm-mcp/contentdm-mcp#diagnostics

## Score history

- 2026-10-18: 79
- 2026-10-17: 79
- 2026-10-16: 79
- 2026-10-15: 79
- 2026-10-14: 79
- 2026-10-13: 50
- 2026-10-12: 50
- 2026-10-08: 35

## Common questions

### What is the CONTENTdm State Archives MCP server?

CONTENTdm State Archives is an MCP server listed in the public MCP registry as io.github.ianderso/contentdm-mcp. Search US state archives' record images on CONTENTdm; read items, pages and transcripts. This page covers its PyPI package (contentdm-mcp).

### Is the CONTENTdm State Archives MCP server safe to use?

CONTENTdm State Archives scores 79 out of 100 on VerifyMCP. We found no known CVEs affecting it as of 18 October 2026. Its build provenance is signed and verified. That is a record of what we were able to check automatically, not an endorsement. The category breakdown on this page shows every signal behind the number, including the ones we could not confirm.

### What tools does the CONTENTdm State Archives MCP server expose?

CONTENTdm State Archives exposes 8 tools: list_instances, list_collections, get_collection, search, get_item, and 3 more. Their descriptions and schemas cost roughly 1,607 tokens of context every time the server is loaded.

### Is the CONTENTdm State Archives MCP server still maintained?

CONTENTdm State Archives is still listed as active in the MCP registry. We last reached this channel on 18 October 2026. Those dates come from our own scans of the registry and the channel itself, not from anything the publisher announced.

### What licence is the CONTENTdm State Archives MCP server under?

CONTENTdm State Archives declares the MIT licence, which is OSI-approved. That covers the source only, and says nothing about the cost of any service it calls.

## Links

- PyPI project: https://pypi.org/project/contentdm-mcp/
- Socket report: https://socket.dev/pypi/package/contentdm-mcp
- Repository: https://github.com/ianderso/contentdm-mcp
- Changelog RSS feed: https://verifymcp.io/servers/ianderso-contentdm-mcp/contentdm-mcp.xml
- Changelog JSON feed: https://verifymcp.io/servers/ianderso-contentdm-mcp/contentdm-mcp.json
- HTML version of this page: https://verifymcp.io/servers/ianderso-contentdm-mcp/contentdm-mcp