Not affiliated with, endorsed by, or sponsored by Beeper or its affiliates. Use at your own risk.

A read-only CLI for AI agents (and humans) to browse, open, and search your local Beeper chat history.

• Read-only access to your local Beeper SQLite database
• Thread listing, detail views, and message browsing
• Built-in full-text search using Beeper's FTS index (with LIKE fallback)
• Optional context windows around search matches
• Optional DM name resolution via platform bridge databases
• JSON output for easy agent integration

• Go 1.22+
• SQLite driver (CGO). On macOS this is typically already available.

# from the repo
cd /path/to/beeper-cli

go build ./cmd/beeper-cli

By default the CLI looks for:

• ~/Library/Application Support/BeeperTexts/index.db
• ~/Library/Application Support/Beeper/index.db

Override with:

• --db /path/to/index.db
• BEEPER_DB=/path/to/index.db

Disable bridge DB lookups with:

• --no-bridge

beeper-cli --help

beeper-cli threads list --days 7 --limit 50
beeper-cli threads show --id "!abc123:beeper.local"

beeper-cli messages list --thread "!abc123:beeper.local" --limit 50

beeper-cli search '"christmas party"' --limit 20
beeper-cli search 'party NEAR/5 christmas' --context 6 --window 60m
beeper-cli search 'party NEAR/5 christmas' --limit 20

beeper-cli threads list --json
beeper-cli search 'invoice' --json

• threads list — list conversations ordered by last activity
• threads show — show thread metadata and participants
• messages list — read recent messages in a thread
• search — full-text search across messages (FTS5)
• db info — show resolved database path and FTS availability
• version — print the current version

Beeper already ships an FTS5 index (mx_room_messages_fts) populated by triggers. The CLI uses that table directly, so no importer is required for keyword or phrase search. If the table doesn't exist, it falls back to a basic LIKE search on message text.

Examples:

• Phrase search: "christmas party"
• Proximity: party NEAR/5 christmas
• Prefix: christ*

Vector search is not built-in to Beeper's SQLite schema, so it requires an additional index:

Option A — Local embeddings + SQLite table

• Build a small local embedding store (message_id -> vector)
• Use a SQLite extension (e.g., sqlite-vss) for cosine similarity
• Pros: stays local; simple deployments
• Cons: requires extra setup and embedding model

Option B — Local embeddings + separate index file

• Use an on-disk index (e.g., HNSW) stored alongside the DB
• Pros: fast and accurate
• Cons: more code and versioned index management

Option C — External vector store

• Push embeddings to a vector DB or hosted index
• Pros: scalable and fast
• Cons: not fully local, extra infrastructure

• This tool is read-only and does not send messages.
• The underlying schema may change as Beeper evolves. If that happens, the queries may need updates.

MIT