Doc2md#

Overview#

This skill provides a packaged doc2md CLI for converting local PDF, DOCX, and PPTX files into Markdown through the doc2md platform API.

Use this skill when the user needs a ready-to-run command-line workflow instead of writing custom conversion code.

Bundled Files#

• Windows binary: scripts/doc2md-cli.exe
• Linux binary: scripts/doc2md-cli

Both binaries are statically linked and do not require an extra runtime.

Prerequisites#

• A valid doc2md bearer token
• Network access to the doc2md API service
• Input files in PDF, DOCX, or PPTX format

Configuration can be provided with either:

• Environment variables: DOC2MD_API_BASE_URL, DOC2MD_BEARER_TOKEN
• Config file: ~/.doc2md/config.json

Environment variables take precedence over the config file.

Usage#

Windows PowerShell#

powershell

$env:DOC2MD_BEARER_TOKEN = 'your-jwt'
$env:DOC2MD_API_BASE_URL = 'http://192.168.99.85:5173'

.\scripts\doc2md-cli.exe -output-dir .\converted .\document.pdf

Linux#

bash

export DOC2MD_BEARER_TOKEN='your-jwt'
export DOC2MD_API_BASE_URL='http://192.168.99.85:5173'

./scripts/doc2md-cli -output-dir ./converted ./document.pdf

Common Commands#

bash

# Convert one folder recursively
./scripts/doc2md-cli -output-dir ./converted ./docs/

# Convert multiple inputs in parallel
./scripts/doc2md-cli -output-dir ./converted -concurrency 4 ./doc1.pdf ./doc2.docx ./folder/

Flags#

FlagDefaultDescription-output-dirrequiredDirectory that receives extracted results-converter``mineruBackend converter: mineru or marker``-recursive``trueScan directories recursively-keep-zip``falseKeep result_clean.zip after extraction-overwrite``trueOverwrite existing output directories-concurrency``1Number of files processed in parallel-poll-interval``5sJob polling interval-job-timeout``24hPer-file timeout-http-timeout``2mPer-request HTTP timeout

Output Behavior#

• Each input document is written to its own subdirectory under -output-dir
• The CLI prints timestamped progress and per-file status to stdout
• Ctrl+C cancels remaining work gracefully
• The process exits non-zero when any conversion fails

Config File Example#

json

{
  "apiBaseUrl": "http://192.168.99.85:5173",
  "bearerToken": "your-jwt"
}

Save this file as ~/.doc2md/config.json when environment variables are not convenient.

Troubleshooting#

• Authentication failure: verify DOC2MD_BEARER_TOKEN
• Connection failure: verify DOC2MD_API_BASE_URL and service reachability
• Empty output directory: confirm the input file type is supported and the API job completed successfully
• Existing output overwritten unexpectedly: pass a safer output directory or adjust -overwrite