> For the complete documentation index, see [llms.txt](https://icij.gitbook.io/datashare/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://icij.gitbook.io/datashare/concepts/cli-reference.md).

# CLI reference

Datashare's command-line interface is organized as subcommands. Run `datashare --help` for an overview, or `datashare COMMAND --help` for help on a specific subcommand.

Running `datashare` with no arguments is equivalent to `datashare app start` — it starts the web application in the default mode.

## Subcommands

| Subcommand               | Description                                         |
| ------------------------ | --------------------------------------------------- |
| `app start`              | Start the Datashare web application server          |
| `stage run`              | Run one or more document processing pipeline stages |
| `worker run`             | Start an async task worker daemon                   |
| `plugin list`            | List available plugins                              |
| `plugin install <id>`    | Install a plugin by id, URL, or file path           |
| `plugin delete <id>`     | Remove an installed plugin                          |
| `extension list`         | List available extensions                           |
| `extension install <id>` | Install an extension by id, URL, or file path       |
| `extension delete <id>`  | Remove an installed extension                       |
| `api-key get`            | Get the current user's API key                      |
| `api-key create`         | Create an API key                                   |
| `api-key delete`         | Delete an API key                                   |

## `app start`

Start the Datashare web application. The default mode is `LOCAL` (single-user, embedded services). Running `datashare` with no arguments is a shorthand for `datashare app start`.

```bash
# Local mode (default)
datashare app start

# Server mode
datashare app start --mode SERVER \
  --elasticsearchAddress http://elasticsearch:9200 \
  --redisAddress redis://redis:6379

# Embedded mode (all services in one JVM)
datashare app start --mode EMBEDDED
```

## `stage run`

Run one or more document processing [stages](/datashare/concepts/cli-stages.md). Requires `--stages`.

```bash
# Scan and index documents
datashare stage run --stages SCAN,INDEX \
  --dataDir /path/to/documents \
  --elasticsearchAddress http://elasticsearch:9200

# Extract named entities
datashare stage run --stages NLP \
  --nlpPipeline CORENLP \
  --elasticsearchAddress http://elasticsearch:9200

# Full pipeline
datashare stage run --stages SCAN,INDEX,NLP \
  --dataDir /path/to/documents \
  --elasticsearchAddress http://elasticsearch:9200
```

## `worker run`

Start an async task worker that processes background tasks (batch searches, batch downloads, indexing, NER extraction).

```bash
datashare worker run \
  --redisAddress redis://redis:6379 \
  --busType REDIS
```

## `plugin` and `extension`

```bash
# List available plugins
datashare plugin list

# Install a plugin
datashare plugin install datashare-plugin-ner-corenlp

# Remove a plugin
datashare plugin delete datashare-plugin-ner-corenlp

# List available extensions
datashare extension list

# Install an extension
datashare extension install datashare-extension-nlp-opennlp

# Remove an extension
datashare extension delete datashare-extension-nlp-opennlp
```

## Global flags

These flags can be placed before any subcommand and apply to all subcommands.

| Flag                     | Default                  | Description                                  |
| ------------------------ | ------------------------ | -------------------------------------------- |
| `-d, --dataDir`          | `~/Datashare`            | Document source directory                    |
| `-P, --defaultProject`   | `local-datashare`        | Default project name                         |
| `-s, --settings`         | `./dist/datashare.conf`  | Path to settings file                        |
| `--logLevel`             | `INFO`                   | Log level (`DEBUG`, `INFO`, `WARN`, `ERROR`) |
| `--elasticsearchAddress` | `http://localhost:9200`  | Elasticsearch URL                            |
| `--redisAddress`         | `redis://localhost:6379` | Redis URL                                    |
| `--busType`              | `MEMORY`                 | Data bus type (`MEMORY`, `REDIS`, `AMQP`)    |
| `--queueType`            | `MEMORY`                 | Queue type (`MEMORY`, `REDIS`, `AMQP`)       |
| `--no-color`             | —                        | Disable colored output                       |
| `-h, --help`             | —                        | Show help and exit                           |
| `-V, --version`          | —                        | Print version and exit                       |

## Backward compatibility

The legacy flag-based syntax (`--mode`, `--stage`, `-m CLI`, etc.) is still accepted for backward compatibility. New scripts should use the subcommand syntax.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://icij.gitbook.io/datashare/concepts/cli-reference.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.