Doc2md

Why it's ranked here

This is a focused conversion component with practical operational choices. Its broad format gate, predictable JSON response, container health check, upload limit, multi-architecture builds, and tested failure cases make it credible for private automation systems.

What's good

Byte-based type detection avoids trusting misleading filenames. Spreadsheet output preserves worksheet structure and reports tab names for downstream chunking. Responses also expose input type, byte count, and conversion duration. The container runs as a non-root user and includes a health check.

Tradeoffs

The service deliberately provides no authentication or rate limiting, and it allows cross-origin requests broadly. It buffers each upload in memory before conversion. Image text extraction and audio transcription require an optional external API key, despite those formats passing initial validation.

How to use it well

Use it inside a private container network when an automation, ingestion, or retrieval pipeline needs Markdown from mixed office files. Put an authenticated reverse proxy in front if exposure is necessary. It handles conversion, not access control, traffic management, downstream chunking, indexing, or retrieval.

Observed

License

MIT

Primary language

Python

Packaging

Docker image published through GitHub Container Registry; local installation uses requirements.txt

Interface

HTTP JSON API with multipart file upload, health endpoint, root metadata endpoint, and interactive FastAPI documentation

Platform support

Container builds target linux/amd64 and linux/arm64

Test structure

End-to-end FastAPI tests cover successful conversions and key upload error cases