Personal Relationship Manager (PRM)

A CLI tool for tracking relationships, interactions, and reminders. Privacy-first: all data stays on your machine. No external APIs.

Setup

Install Rust (stable, 1.93+)
Install Ollama for AI features, then pull the default model:
```
ollama pull llama3.2:3b
```

Clone and run:

git clone https://github.com/petros-p/prm.git
cd prm
cargo run

Data is stored in .data/prm.db (SQLite).

Commands

People

Command	Description
`people`	List all people
`add-person [name]`	Add a person (interactive)
`show-person <name>`	Show person details (with recent interactions)
`history <name>`	Show full interaction history
`edit-person <name>`	Edit a person (menu)
`find <query>`	Search people, circles, labels
`archive-person <name>`	Archive a person
`unarchive-person <name>`	Restore archived person
`archived-people`	List archived people

Quick Edits

Command	Description
`edit-name <name>`	Edit person's name
`edit-nickname <name>`	Edit nickname
`edit-birthday <name>`	Edit birthday
`edit-how-we-met <name>`	Edit how you met
`edit-notes <name>`	Edit notes
`edit-location <name>`	Edit location
`edit-labels <name>`	Edit labels
`edit-circles <name>`	Edit circles
`edit-phone <name>`	Edit phone numbers
`edit-email <name>`	Edit email addresses

Circles

Command	Description
`circles`	List all circles
`add-circle [name]`	Create a circle
`show-circle <name>`	Show circle details
`edit-circle <name>`	Edit a circle
`archive-circle <name>`	Archive a circle
`unarchive-circle <name>`	Restore archived circle
`archived-circles`	List archived circles

Labels

Command	Description
`labels`	List all labels
`add-label [name]`	Create a label
`show-label <name>`	Show label details
`edit-label <name>`	Edit a label
`archive-label <name>`	Archive a label
`unarchive-label <name>`	Restore archived label
`archived-labels`	List archived labels

Interactions & Reminders

Command	Description
`log <name>`	Log an interaction (manual)
`ai-log <description>`	Log via AI (natural language, local Ollama)
`voice-log <wav-file>`	Log via voice recording (local Whisper transcription)
`remind`	Show overdue and upcoming reminders
`set-reminder <name>`	Set reminder frequency

Other

Command	Description
`stats`	Show statistics
`help`	Show all commands
`exit` / `quit` / `q`	Exit

AI Features

All AI runs locally — no API keys, no data leaves your machine.

ai-log — Describe an interaction in plain text; Ollama parses it into structured data for review and save.
voice-log — Record a .wav file; Whisper transcribes it locally, then Ollama parses it.

Voice log setup

Download the Whisper model (~148MB) and place it at .data/models/ggml-base.en.bin:

mkdir -p .data/models
curl -L -o .data/models/ggml-base.en.bin \
  https://huggingface.co/ggerganov/whisper.cpp/resolve/main/ggml-base.en.bin

Environment variables

Variable	Default	Description
`OLLAMA_HOST`	`http://localhost:11434`	Ollama server URL
`PRM_MODEL`	`llama3.2:3b`	Ollama model to use
`PRM_WHISPER_MODEL`	`.data/models/ggml-base.en.bin`	Whisper model path

Building

cargo build --release    # Optimized binary at target/release/prm
cargo test               # Run all tests (99 tests)
cargo check              # Fast compile check