# claudia-docs-cli

CLI tool for interacting with Claudia Docs from agents and scripts.

## Installation

### Download Binary
```bash
# Linux/macOS
curl -fsSL https://gitea.danielarroyo.cl/proyectos/claudia-docs-cli/releases/latest/claudia-docs -o /usr/local/bin/claudia-docs
chmod +x /usr/local/bin/claudia-docs
```

### Build from Source
```bash
git clone https://gitea.danielarroyo.cl/proyectos/claudia-docs-cli.git
cd claudia-docs-cli
go build -o claudia-docs ./cmd/claudia-docs
```

## Quick Start

```bash
# Login
./claudia-docs auth login -u admin -p your_password --save

# List projects
./claudia-docs project list

# Create document
./claudia-docs doc create -t "My Document" -c "# Hello" -p <project-id>

# List documents
./claudia-docs doc list --project-id <project-id>
```

## Global Options

| Flag | Description | Default |
|------|-------------|---------|
| `--server` | API server URL | `http://localhost:8000` |
| `--token` | JWT token | From config or `CLAUDIA_TOKEN` env |
| `--output` | Output format: `json`, `text` | `json` |
| `--quiet` | Suppress stdout, only JSON | `false` |
| `--config` | Config file path | `~/.claudia-docs.yaml` |
| `--verbose` | Verbose debug output | `false` |

## Environment Variables

| Variable | Description |
|----------|-------------|
| `CLAUDIA_SERVER` | API server URL |
| `CLAUDIA_TOKEN` | JWT token |
| `CLAUDIA_OUTPUT` | Output format |

## Commands

### Auth

```bash
# Login and save token
./claudia-docs auth login -u <username> -p <password> --save

# Check status
./claudia-docs auth status

# Logout
./claudia-docs auth logout
```

### Projects

```bash
# List all projects
./claudia-docs project list

# Get project details
./claudia-docs project get <project-id>
```

### Documents

```bash
# Create document
./claudia-docs doc create -t "Title" -c "# Content" -p <project-id> [-f <folder-id>]

# List documents in project
./claudia-docs doc list --project-id <project-id> [--limit 20] [--offset 0]

# Get document
./claudia-docs doc get <document-id> [--include-reasoning]

# Update document
./claudia-docs doc update <document-id> [-t "New Title"] [-c "New Content"]

# Delete document
./claudia-docs doc delete <document-id> [--force]
```

### Folders

```bash
# List folders in project
./claudia-docs folder list --project-id <project-id> [--parent-id <folder-id>]

# Create folder
./claudia-docs folder create --project-id <project-id> -n "Folder Name" [--parent-id <folder-id>]
```

### Tags

```bash
# List all tags
./claudia-docs tag list

# Create tag
./claudia-docs tag create -n <name> [--color <hex-color>]

# Add tag to document
./claudia-docs tag add --doc-id <document-id> --tag-id <tag-id>
```

### Search

```bash
# Full-text search
./claudia-docs search -q "query" [--project-id <project-id>] [--tags <tag1,tag2>]
```

### Reasoning

```bash
# Save reasoning metadata
./claudia-docs reasoning save -d <document-id> -t <type> -s '<json-steps>' [--confidence 0.85] [--model <model-name>]

# Types: research, planning, analysis, synthesis
# Steps JSON: '[{"step_id":"1","thought":"...","conclusion":"..."}]'
```

## Output Format

### JSON (default)
```json
{
  "success": true,
  "data": {...},
  "error": null,
  "meta": {
    "command": "doc create",
    "duration_ms": 45,
    "server": "http://localhost:8000"
  }
}
```

### Text
```bash
./claudia-docs --output text project list
```

## Configuration

Config file: `~/.claudia-docs.yaml`

```yaml
server: http://localhost:8000
token: ""
timeout: 30s
output: json

agents:
  default:
    token: ""
    default_project: ""
```

## Examples

### Full Workflow

```bash
# 1. Login
./claudia-docs auth login -u researcher -p secret --save

# 2. List projects
./claudia-docs project list

# 3. Create document with content
./claudia-docs doc create \
  --title "Research: AI Trends 2026" \
  --content "# AI Trends\n\n..." \
  --project-id "uuid-project"

# 4. Search
./claudia-docs search -q "AI trends"

# 5. Get and update
./claudia-docs doc get <doc-id>
./claudia-docs doc update <doc-id> -t "Updated Title"
```

### Script Integration

```bash
#!/bin/bash
export CLAUDIA_SERVER=http://localhost:8000
export CLAUDIA_TOKEN=$(./claudia-docs auth login -u agent -p pass --output json | jq -r '.data.token')

# Create document
DOC_ID=$(./claudia-docs doc create -t "Report" -c "# Content" -p $PROJECT_ID --output json | jq -r '.data.id')

# Save reasoning
./claudia-docs reasoning save -d $DOC_ID -t research -s '[{"step_id":"1","thought":"Analysis...","conclusion":"Result"}]'
```

## Help

```bash
./claudia-docs --help
./claudia-docs auth --help
./claudia-docs doc --help
```





---

# En tu máquina local
git tag v1.0.0
git push origin v1.0.0
```

---

### Secret necesario

En **Settings → Secrets → Actions** de tu repo:

| Secret | Cómo obtenerlo |
|---|---|
| `RELEASE_TOKEN` | Gitea → Settings → Applications → Generate Token (con permiso `write:repository`) |

---

### Estructura esperada del proyecto
```
mi-proyecto/
├── cmd/
│   └── main.go        ← entrypoint
├── go.mod
├── go.sum
└── .gitea/
    └── workflows/
        └── release.yml

# Si main.go está en la raíz
-o "dist/${BINARY_NAME}" .

# Si está en otro lugar
-o "dist/${BINARY_NAME}" ./internal/cmd/server.go
```

---

### Resultado en Gitea

Una vez que hagas push del tag, en **Releases** del repo verás:
```
Release v1.0.0
├── app-linux-amd64.tar.gz
├── app-linux-arm64.tar.gz
├── app-darwin-amd64.tar.gz
├── app-darwin-arm64.tar.gz
├── app-windows-amd64.zip
└── checksums.txt