---
title: "Voice notes as AI context: use cases for Signal Recorder SR-7"
source: https://signalrecorder.app/docs/use-cases
---

# Voice notes as AI context: use cases for Signal Recorder SR-7

## From voice to a searchable archive

The simplest use of Signal Recorder SR-7 is also the one that pays off most: record freely and let the app do the rest. Each recording is transcribed on your device and, with Apple Intelligence, gets a title and a one-sentence summary. Over weeks and months that adds up to an archive of what you were thinking, searchable by every word.

### A daily voice journal

Record a few minutes at the end of each day. Don't worry about structure; just talk. SR-7 transcribes it, writes a title such as "Sprint retrospective thoughts and hiring update" and a summary, and adds it to your library. Weeks later, search for "hiring" or "sprint" and find what you said.

### Your recap after a meeting or an interview

SR-7 records your own voice; it does not join calls or label speakers. The habit that works is a recap: walk out of the meeting or put down the phone, and say what was decided, what is open and what you think. The transcript becomes your searchable record, and the summary gives you the gist without listening again.

For research, group the debriefs in a project. After the fifth or sixth, search across all of them for the pattern.

### Thinking out loud

Some ideas only take shape when you say them. Talk a problem through on a walk, and the transcript is waiting when you sit down. With iCloud sync switched on, a recording made on your iPhone shows up on your Mac, transcribed.

## Working with AI tools through the MCP server

On the Mac, SR-7 runs a local MCP server, so an AI client such as Claude Code can read and search your archive. Your recordings become a source the agent can work from. The [MCP server guide](https://signalrecorder.app/docs/mcp-server) covers setup and the 20 tools; these are the kinds of things to ask.

### Search your recordings

"What did I say about the API redesign last week?" The client uses `search`, then `get_recording` on what it finds, and answers from your own words.

### Pull out what you said you would do

"Go through today's recordings and list anything I committed to." The client lists the day's recordings with `list_recordings` and reads each transcript.

### Summarize a project

"Give me an overview of everything in the Product Launch project." An agent working from 20 recordings in one project gives a far better overview than one working from your whole library.

### Draft from your voice notes

"I recorded some thoughts about on-device AI. Turn them into a blog post draft." The agent searches for the recordings, reads them and drafts from your phrasing. Your voice notes become first drafts.

### Tidy as you go

"Move this week's journal entries into the Journal project," or "clean up the transcript of my last recording." The client can move recordings, edit text and create notes, within the tools you have switched on.

## A markdown folder for Obsidian and other tools

On the Mac, SR-7 can keep every recording as a markdown file. In Settings, under Library, pick a location and SR-7 creates an SR-7 Library folder there, with a folder per project. Each file starts with YAML front matter (`uuid`, `created`, `duration` when the recording has one, `audio` for recordings, and `favorite` when you have starred it), then the title as a heading, the summary as a quote and the transcript.

### Build a voice-first knowledge base

Put the library folder inside your Obsidian vault and your voice notes sit alongside your written notes, searchable with everything else. SR-7 manages the folder: it rewrites files when recordings change and removes files it did not write. Edit in SR-7, read in Obsidian, and keep your own notes outside the SR-7 Library folder.

### Feed recordings into other tools

The files work as input for anything that reads text: search indexes, static site generators, scripts. The front matter makes them easy to filter, and the `uuid` stays the same when a title changes.

## Tips

- **Record short and often.** Five one-minute recordings are more useful than one rambling ten-minute recording. Shorter recordings get sharper titles and summaries.
- **Use projects to group context.** An agent summarizing a project gives a far better overview than one summarizing your whole library.
- **Star what you come back to.** Favorites cut across projects: use them for key decisions and ideas you keep returning to.
- **Leave automatic transcription on.** With **Transcribe automatically on import** switched on, as it is to begin with, every import is searchable soon after it lands.