# file2markdown

> Convert docs and web pages to clean Markdown: PDF, DOCX, XLSX, EPUB, scans, URLs. Batch of 10.

- name: ai.file2markdown/file2markdown
- version: 1.1.0
- connection class: R0 (autonomous)
- trust: 96/100
- quality: 83/100 (good)
- owner: not claimed
- descriptor: https://api.protogrid.dev/v1/servers/ai.file2markdown%2Ffile2markdown
- tools: https://api.protogrid.dev/v1/servers/ai.file2markdown%2Ffile2markdown/tools

**An agent can connect right now, no human step.** Remote endpoint, no authentication, reachable on the last probe.

## Connection (mcpServers)

```json
{
  "mcpServers": {
    "file2markdown": {
      "type": "http",
      "url": "https://mcp.file2markdown.ai/mcp",
      "headers": {
        "Authorization": "Bearer ${FILE2MARKDOWN_TOKEN}"
      }
    }
  }
}
```

Placeholders to fill: `${FILE2MARKDOWN_TOKEN}`.

## Trust

- hygiene: 100
- liveness: 100
- freshness: 100
- provenance: 85
- drivers: +repository +dns-namespace +website +reachable +uptime-100% +updated-8d-ago
- Derived from observable signals (official registry feed and our own credential-free probes); no code audit performed.

## Quality 83/100 (good)

### protocol: 88 (weight 25)

- pass `protocol.modern`: supports 2026-07-28 (server/discover)
- pass `protocol.stateless`: answers without a session
- pass `protocol.transport`: streamable HTTP
- warn `protocol.list_ttl`: tools/list declares ttlMs 0, so clients re-fetch it on every use

### authorization: n/a (weight 15)

- n/a `auth.prm`, `auth.as_metadata`, `auth.cimd`: no remote uses OAuth
- n/a `auth.secret_in_url`: no templated URL

### tool hygiene: 71 (weight 30)

- pass `tools.descriptions`: every tool has a description
- pass `tools.description_length`: descriptions are concise
- pass `tools.schemas`: every tool has an object input schema
- fail `tools.annotations`: no tool declares readOnlyHint or destructiveHint, so clients cannot tell safe calls from risky ones
- fail `tools.directory_hints`: no tool declares all four of readOnlyHint, destructiveHint, idempotentHint and openWorldHint; missing: usage_status (readOnlyHint, destructiveHint, idempotentHint, openWorldHint); convert_url (readOnlyHint, destructiveHint, idempotentHint, openWorldHint); convert_base64 (readOnlyHint, destructiveHint, idempotentHint, openWorldHint); …
- pass `tools.token_cost`: about 840 tokens to load every tool
- pass `tools.api_dump`: tools are not a one-to-one API dump

### stability: 100 (weight 15)

- pass `stability.changes`: 1 tool changes in 30 days
- pass `stability.rug_pull`: no tool changed its meaning under the same name

### dependencies: n/a (weight 15)

- n/a `deps.known_vulns`, `deps.mcp_sdk_version`, `deps.resolvable`: no npm or PyPI package

- tools: 5, about 840 tokens to load them all
- tool set hash: 1a280bca261a9618575a3dea7ea1c911
- tools last changed: 2026-10-13T01:26:45.223Z
- badge: [![protogrid quality](https://protogrid.dev/badge/ai.file2markdown/file2markdown.svg)](https://protogrid.dev/servers/ai.file2markdown/file2markdown)
- Checks run on what our credential-free, read-only probes observe; tools are never called and no code audit is performed.

### Tool changes

- 2026-10-13 `convert_batch` added
  - Convert 2-10 documents to Markdown in one call (Pro API key required). Each item is either {"url": ..., "filename_hint"?: ...} or {"content_base64": ..., "filename": ...}; types may be mixed. Items are converted in parallel and isolated: one bad URL or file never fails the batch — its slot returns ok:false with the error. The combined markdown shares one output budget, split evenly across items (per-item `truncated` flags apply). Set include_markdown=false to get only metadata + quality per item (health-check a folder without flooding your context). Anonymous callers: use convert_url / convert_base64 individually.

Full history: https://api.protogrid.dev/v1/servers/ai.file2markdown%2Ffile2markdown/changes

## Remotes

- https://mcp.file2markdown.ai/mcp (streamable-http, auth none, reachable true, uptime30d 1)

## Packages

_none_

## Tools (5)

- `convert_base64`: Convert base64-encoded file contents to Markdown.

    filename is required and drives format detection (e.g. "report.docx").
    Intended for programmatic MCP clients; for anything with a URL, prefer
    convert_url.

    quality_check: "basic" (default) attaches local output-health checks in
    the `quality` field; "ai" additionally sends selected excerpts of the
    converted text to TypeSafe for narrow yes/no judgments. See convert_url.
    
- `convert_batch`: Convert 2-10 documents to Markdown in one call (Pro API key required).

    Each item is either {"url": ..., "filename_hint"?: ...} or
    {"content_base64": ..., "filename": ...}; types may be mixed. Items are
    converted in parallel and isolated: one bad URL or file never fails the
    batch — its slot returns ok:false with the error. The combined markdown
    shares one output budget, split evenly across items (per-item `truncated`
    flags apply). Set include_markdown=false to get only metadata + quality
    per item (health-check a folder without flooding your context).

    Anonymous callers: use convert_url / convert_base64 individually.
    
- `convert_url`: Fetch a URL (web page, PDF, Office doc, and more) and convert it to Markdown.

    Works on publicly accessible URLs. Web pages are converted from their
    served HTML; JavaScript-rendered content may be incomplete. Optional
    filename_hint (e.g. "report.pdf") helps format detection when the URL
    has no file extension.

    The result carries a `quality` object (output-health check of the returned
    Markdown: status review / no_issues_detected / not_evaluated, coverage and
    a list of issues with line locations). quality_check="basic" (default)
    runs only local deterministic checks. quality_check="ai" additionally
    sends selected excerpts of the CONVERTED text to TypeSafe for narrow
    yes/no judgments (garbled text, hard-wrapped prose, broken tables); it is
    off unless requested. No check compares against the source document.
    
- `list_supported_formats`: List supported input formats, per-tier limits, and which features need Pro.

    Single source of truth — do not claim capabilities beyond what this
    returns.
    
- `usage_status`: Check your tier and remaining conversions before burning a call.