> ## Documentation Index
> Fetch the complete documentation index at: https://docs.augent.app/llms.txt
> Use this file to discover all available pages before exploring further.

# Workflows & Outputs

> Route Augent output to notes apps, automate pipelines, and export in any format.

## Output Destinations

Augent pipelines audio from URL to structured notes — download, transcribe, analyze, and format in a single call. The client controls where those results land.

### Obsidian

Augent has a dedicated Obsidian integration with real-time sync. See [Obsidian Setup](/guides/obsidian-setup).

### Notion

Tell your agent to format output as a Notion-compatible page and save it somewhere accessible.

```
Take notes from this podcast and save them as a markdown file I can import to Notion:
https://youtube.com/watch?v=xxx
```

Augent's `take_notes` tool saves `.txt` files formatted with headers, bullets, callouts, and tables. Notion imports markdown directly — drag the file into a Notion page or use **Import → Markdown** in the sidebar.

For structured data (search results, transcriptions), use the `output` parameter to export as `.csv` or `.xlsx`, then import into a Notion database.

### Apple Notes

Apple Notes doesn't support markdown, but it handles plain text and basic formatting.

```
Take notes from this interview and save a clean summary to my Desktop:
https://youtube.com/watch?v=xxx
```

The `.txt` output from `take_notes` opens in any text editor. Copy and paste into Apple Notes, or set up a Shortcut to automate it.

For bulk workflows, export transcriptions and search results to `.csv` or `.xlsx` using the `output` parameter and import into Apple Notes via the share sheet.

### Logseq

Logseq reads markdown from its graph directory. Point Augent's output there.

```
Take notes from this lecture and save to ~/Documents/LogseqGraph/pages/:
https://youtube.com/watch?v=xxx
```

Use the `output_dir` parameter on `take_notes` to save directly into your Logseq graph's `pages/` folder. Logseq picks up new files automatically. The `.txt` format with markdown headers and bullets renders natively in Logseq.

***

## Automation

### n8n

One node triggers Augent's full pipeline — download, transcription, semantic search, speaker identification, and structured notes. No stitching together separate APIs.

1. Add an **Execute Command** node to your workflow
2. Run a single Augent command:

```bash theme={null}
augent notes "https://youtube.com/watch?v=xxx" --output ~/pipeline/notes.txt
```

3. Route the output to the next node in your workflow

For granular control, individual tools are also available:

```bash theme={null}
augent transcribe /path/to/audio.mp3 --format json --output /path/to/output.json
augent search /path/to/audio.mp3 "open source,self-hosted" --format csv --output /path/to/results.csv
```

Combine with triggers (new file in folder, webhook, RSS feed) to fully automate content pipelines. Example: new podcast episode drops → Augent handles the full pipeline — download, transcribe, analyze, and format notes — results land in Notion or Obsidian.

### Make

Add Augent as a node in any Make scenario. One trigger kicks off the full audio-to-notes pipeline — the same end-to-end processing available in n8n.

1. Add a **Shell** module to your scenario
2. Run Augent the same way as n8n above
3. Route the structured output to any of Make's 1,000+ app connectors

***

## Export Formats

Every search and transcription tool supports the `output` parameter for saving results to disk:

| Format   | Extension | Use case                                                       |
| -------- | --------- | -------------------------------------------------------------- |
| CSV      | `.csv`    | Spreadsheets, databases, automation pipelines                  |
| XLSX     | `.xlsx`   | Styled spreadsheets with bold headers and formatted timestamps |
| JSON     | `.json`   | APIs, scripts, programmatic access                             |
| SRT      | `.srt`    | Subtitles for video editors                                    |
| VTT      | `.vtt`    | Web video players                                              |
| Markdown | `.md`     | Documentation, notes apps                                      |

```json theme={null}
{
  "audio_path": "/path/to/audio.mp3",
  "keywords": ["launch date", "pricing"],
  "output": "~/Desktop/results.xlsx"
}
```

The `output` parameter works on `transcribe_audio`, `search_audio`, `deep_search`, `search_proximity`, and `search_memory`.