Aspose.PDF Cloud CLI + MCP Tools v1.1.0

Aspose.PDF Cloud CLI exposes Aspose PDF Cloud storage and PDF operations through both a command-line interface and an MCP server.

Version 1.1.0 adds PDF attachment listing and extraction, upgrades the MCP server to the MCP Python SDK 2.x, and adds Git-based package versioning and release automation. See CHANGELOG.md for the release notes.

How to Get Your Aspose.PDF Cloud Credentials

1. Go to the Dashboard.

2. After logging in, navigate to the Applications section in the main menu.

3. Select the first (default) application or create a new one.

4. Copy the Client ID and Client Secret for later use.

Related MCP server: pdf-mcp

Installation

Requires Python 3.11 or later. Version 1.1.0 uses Aspose.PDF Cloud SDK 26.x and MCP Python SDK 2.x; pip installs these dependencies automatically.

Install from PyPI:

python -m pip install aspose-pdf-cloud-mcp

Check that the CLI is available:

aspose-pdf-cloud-cli --help
apdf --help

For local development from a checkout:

python -m pip install -e ".[dev]"

Configuration

Set credentials before running CLI commands or the MCP server. The easiest local setup path is:

apdf auth login
apdf auth status
apdf auth test

You can also set credentials directly in your shell:

$env:ASPOSE_CLIENT_ID = "your-client-id"
$env:ASPOSE_CLIENT_SECRET = "your-client-secret"

Optional environment variables:

• ASPOSE_STORAGE_NAME: default Aspose storage name.

• ASPOSE_BASE_URL: alternate Aspose PDF Cloud base URL.

• ASPOSE_SELF_HOST: set to 1, true, yes, or on for self-hosted use.

You can also create a local .env file. See .env.example for a template. Environment variables take precedence over .env values.

CLI Usage

The canonical executable is aspose-pdf-cloud-cli. A shorter apdf alias is also installed for convenience.

aspose-pdf-cloud-cli storage list /
aspose-pdf-cloud-cli auth status
aspose-pdf-cloud-cli auth test
aspose-pdf-cloud-cli storage upload .\sample.pdf /sample.pdf
aspose-pdf-cloud-cli storage download /sample.pdf .\sample.pdf
aspose-pdf-cloud-cli storage download /sample.pdf .\sample.pdf --overwrite
aspose-pdf-cloud-cli pdf merge merged.pdf /a.pdf /b.pdf /c.pdf
aspose-pdf-cloud-cli pdf merge merged.pdf --from-folder /contracts
aspose-pdf-cloud-cli pdf split sample.pdf
aspose-pdf-cloud-cli pdf split sample.pdf --ranges 1-3,4,5-8
aspose-pdf-cloud-cli pdf pdfa-versions
aspose-pdf-cloud-cli pdf convert-pdfa sample.pdf /archive/sample-pdfa.pdf --pdfa-version PDF/A-1B
aspose-pdf-cloud-cli pdf convert-pdfa sample.pdf /archive/sample-pdfa.pdf --type PDFA3B
aspose-pdf-cloud-cli pdf extract-text sample.pdf --output sample.txt
aspose-pdf-cloud-cli pdf extract-text sample.pdf --output sample.txt --overwrite
aspose-pdf-cloud-cli pdf extract-tables sample.pdf --output tables.json
aspose-pdf-cloud-cli pdf extract-tables sample.pdf --pages 1,3,4-7,10 --output tables.json
aspose-pdf-cloud-cli pdf list-images sample.pdf
aspose-pdf-cloud-cli pdf extract-images sample.pdf /extracted-images
aspose-pdf-cloud-cli pdf extract-images sample.pdf /extracted-images --pages 1,3,7-10
aspose-pdf-cloud-cli pdf extract-image sample.pdf 1 1 /extracted-images
aspose-pdf-cloud-cli mcp serve

MCP Usage

After installing the package, configure your MCP client to launch:

aspose-pdf-cloud-cli mcp serve

Codex MCP configuration examples are available in docs/mcp-config.md, with a copy-pasteable helper at examples/codex-mcp-config.toml.

Codex Prompts

Sample prompts for using this MCP server with Codex are in docs/codex-prompts.md.

Install as an Agent Skill

The package includes an aspose-pdf-cloud-mcp skill for Codex and Claude Code. Install the package first, then install the bundled skill:

python -m pip install aspose-pdf-cloud-mcp
aspose-pdf-cloud-cli skill install codex
aspose-pdf-cloud-cli skill install claude-code

For Claude Code project-local installation:

aspose-pdf-cloud-cli skill install claude-code --project

Use --force to replace an existing installed copy. More details are in docs/agent-skill.md.

Extract PDF attachments

The PDF must already be in Aspose storage. List embedded files, download one by its one-based index, or download all of them:

apdf pdf list-attachments sample.pdf
apdf pdf extract-attachment sample.pdf 1 ./report.txt
apdf pdf extract-attachments sample.pdf ./attachments

All three commands support --folder and --storage. Extraction writes local files and refuses existing outputs unless --overwrite is specified. Bulk extraction prefixes filenames with their index (for example 1-report.txt), preserving duplicate names safely. Unsafe embedded filenames are rejected before downloads. Empty PDFs produce an empty list. Each local file is written atomically; a later download failure can leave earlier completed files.

Equivalent MCP tools are list_attachments(name, folder, storage), extract_attachment(name, index, local_path, folder, storage, overwrite), and extract_attachments(name, output_dir, folder, storage, overwrite). Listing returns attachments with metadata and index; single extraction returns local_path; bulk extraction returns attachments, output_dir, and files. These tools use the standard MCP success/error envelope.