---
title: "LLMind CLI reference — commands, flags, exit codes | LLMind"
description: "Complete reference for the LLMind command-line tool. Commands: enrich, verify, inspect, sign. Flags, exit codes, and usage examples."
url: https://llmind.org/docs/cli/
source_format: html
---
# LLMind CLI reference

Published 2026-04-22 · 6 min read

Complete reference for the `llmind` command-line tool. Learn all commands, flags, and exit codes.

## Commands

| Command | Purpose | Example |
| --- | --- | --- |
| `llmind enrich` | Analyze a file and write a semantic layer inside it as XMP metadata. | `llmind enrich document.pdf` |
| `llmind verify` | Check the cryptographic signatures on an enriched file to confirm the semantic layer has not been tampered with. | `llmind verify document.pdf` |
| `llmind inspect` | Display the semantic layer embedded in a file—description, entities, structure, and signatures. | `llmind inspect document.pdf` |
| `llmind sign` | Add or refresh cryptographic signatures on an existing semantic layer. | `llmind sign document.pdf` |
| `llmind --version` | Print the installed version. | `llmind --version` |
| `llmind --help` | Show built-in help and available subcommands. | `llmind --help` |

## Common flags

`--recursive`

Apply the command to all matching files in a directory tree. Example: `llmind enrich --recursive ~/documents`

`--sign`

Sign the semantic layer with a cryptographic signature during enrichment (or when re-signing). Verifies that the layer has not been modified since creation.

`--layer <name>`

Specify a custom XMP namespace for the semantic layer. Default is `http://llmind.org/semantic/`. Useful for versioning or multi-layer enrichment.

`--from-stdin`

Read file content from standard input instead of a file path. Useful for piping or processing streams.

`--ocr-provider <name>`

Specify the OCR provider for image and scanned PDF enrichment. Example: `--ocr-provider openai`. Defaults to built-in provider.

`--lineage-source <url>`

Embed a URL indicating where the file came from. Useful for tracking provenance in datasets and training pipelines.

`--lineage-license <spdx>`

Embed an SPDX license identifier indicating how the file may be used. Example: `--lineage-license CC-BY-4.0`.

## Exit codes

| Code | Meaning |
| --- | --- |
| `0` | Success. Command completed without errors. |
| `1` | General error. File not found, invalid arguments, or I/O failure. |
| `2` | Signature verification failed. The semantic layer has been tampered with or the signature is invalid. |
| `3` | Unsupported file format. The file type is not supported by LLMind. |

## Examples

### Enrich a single file

```
llmind enrich report.pdf
```

### Enrich an entire directory recursively

```
llmind enrich --recursive ~/documents
```

### Enrich and sign

```
llmind enrich --sign report.pdf
```

### Verify an enriched file has not been tampered with

```
llmind verify report.pdf
```

### Inspect the semantic layer

```
llmind inspect report.pdf
```

### Enrich with provenance metadata

```
llmind enrich --lineage-source https://example.com/dataset.pdf --lineage-license CC-BY-4.0 report.pdf
```

## Next steps

-   [Quickstart](https://llmind.org/docs/quickstart/) — get started in 5 minutes
-   [Install](https://llmind.org/docs/install/) — detailed installation guide
-   [Dataset provenance](https://llmind.org/use-cases/dataset-provenance/) — use lineage flags to track training data
-   [Product](https://llmind.org/product/) — overview of LLMind
-   [MCP server](https://llmind.org/mcp/) — integrate with agents