Beeper

CRITICAL: TOKENS PAY RENT

Every output token must produce actionable value. Violations:

1. NO PASSIVE SUMMARIES - Regurgitating conversation content without action items, code, or artifacts is FORBIDDEN
2. NO AGREEMENT WITHOUT IMPLEMENTATION - "I agree with X" must be followed by code/file/commit implementing X
3. NO RHETORICAL QUESTIONS - Ask only when you cannot proceed without the answer
4. NO PRAISE/VALIDATION - Skip "great question" / "you're right" - proceed to work

When reviewing message history:

• Extract ACTION ITEMS → create files, send messages, write code
• Extract DECISIONS → update configs, create artifacts documenting the decision
• Extract BLOCKERS → file issues, send follow-up messages
• NEVER just summarize what was discussed

Enforcement: If output contains summary without artifact, STOP and create the artifact first.

Beeper Unified Messaging Skill

Access all messaging networks through three access tiers with increasing depth.

Three Access Tiers

Tier 1: Desktop API (MCP + HTTP) — real-time chat; send text + attachments
Tier 2: beeper-cli (Auth)   — paginated history, chat type metadata, contacts
Tier 3: SQLite→DuckDB (Archive) — full offline archive, cross-platform analytics

When to Use Which Tier

Tier 1: Desktop API (MCP + HTTP)

# Search for a chat
mcp__beeper__search_chats query="contact name"

# Get chat metadata
mcp__beeper__get_chat chatID="..."

# List messages from a chat
mcp__beeper__list_messages chatID="..."

# Search message content (literal word match; limit <= 20)
mcp__beeper__search_messages query="keyword" limit=20

# Send a text message
mcp__beeper__send_message chatID="..." text="Hello!"

# Prefill a draft with an attachment (manual send fallback)
mcp__beeper__focus_app chatID="..." draftText="..." draftAttachmentPath="/path/to/file"

MCP Tools

Send Attachments (Programmatic)

MCP send_message is text-only. To send files, use Beeper Desktop API:

1. POST /v1/assets/upload (multipart) → returns uploadID
2. POST /v1/chats/{chatID}/messages with JSON body containing attachment.uploadID

Wrapper script (recommended): scripts/beeper_send_file.sh

skills/beeper/scripts/beeper_send_file.sh '<chat_id>' /path/to/file 'optional text'

If curl to http://localhost:23373 fails in a sandbox, rerun with escalated permissions.

Tier 2: beeper-cli (Authenticated CLI)

Requires fnox for auth. Assumes beeper-cli is available in PATH.

# Verify beeper-cli is available
command -v beeper-cli

# Auth pattern — secret never exposed to context
BEEPER_ACCESS_TOKEN=$(fnox get BEEPER_ACCESS_TOKEN --age-key-file ~/.age/key.txt) \
  beeper-cli <command> -o json

# List chats (has type: "single" vs "group" — more reliable than MCP)
BEEPER_ACCESS_TOKEN=$(fnox get BEEPER_ACCESS_TOKEN --age-key-file ~/.age/key.txt) \
  beeper-cli chats list -o json

# List messages from a chat
BEEPER_ACCESS_TOKEN=$(fnox get BEEPER_ACCESS_TOKEN --age-key-file ~/.age/key.txt) \
  beeper-cli messages list -o json --chat-id "..."

# List connected accounts (Signal, Telegram, WhatsApp, iMessage)
BEEPER_ACCESS_TOKEN=$(fnox get BEEPER_ACCESS_TOKEN --age-key-file ~/.age/key.txt) \
  beeper-cli accounts list -o json

# Filter by account
BEEPER_ACCESS_TOKEN=$(fnox get BEEPER_ACCESS_TOKEN --age-key-file ~/.age/key.txt) \
  beeper-cli chats list -o json --account-ids whatsapp

# Pagination (cursor-based)
BEEPER_ACCESS_TOKEN=$(fnox get BEEPER_ACCESS_TOKEN --age-key-file ~/.age/key.txt) \
  beeper-cli chats list -o json --cursor "..." --direction before

beeper-cli Caveats

• Pagination bug: cursor can cycle — always deduplicate on chat ID
• No --limit flag: uses cursor pagination only
• iMessage coverage: shows ~22 recent chats; chat.db has 354 DMs
• WhatsApp: separate account, must use --account-ids whatsapp

Connected Accounts

Tier 3: SQLite→DuckDB (Full Archive)

Direct database access for complete offline analysis. DuckDB attaches SQLite databases inline — no ETL needed.

Database Locations

DuckDB Inline Attach Pattern

INSTALL sqlite;
LOAD sqlite;

-- Attach both databases read-only
ATTACH '~/Library/Application Support/BeeperTexts/account.db'
  AS beeper (TYPE sqlite, READ_ONLY);
ATTACH '~/Library/Messages/chat.db'
  AS imessage (TYPE sqlite, READ_ONLY);

-- Now query across both with standard SQL
SELECT * FROM beeper.local_events LIMIT 1;
SELECT * FROM imessage.chat LIMIT 1;

Unified DM Landscape Query

The canonical cross-platform DM analysis query:

INSTALL sqlite;
LOAD sqlite;

ATTACH '~/Library/Application Support/BeeperTexts/account.db'
  AS beeper (TYPE sqlite, READ_ONLY);
ATTACH '~/Library/Messages/chat.db'
  AS imessage (TYPE sqlite, READ_ONLY);

WITH
-- Beeper Signal + Telegram DMs (3+ user messages, ≤2 senders = DM)
beeper_dms AS (
    SELECT
        le.room_id,
        COUNT(*) FILTER (WHERE le.sender = '@zigger:beeper.com') AS my_msg_count,
        COUNT(DISTINCT le.sender) AS sender_count,
        CASE
            WHEN le.room_id LIKE '%.local-signal.%' THEN 'Signal'
            WHEN le.room_id LIKE '%.local-telegram.%' THEN 'Telegram'
            ELSE 'Other'
        END AS network
    FROM beeper.local_events le
    WHERE le.type = 'm.room.message'
    GROUP BY le.room_id
    HAVING COUNT(*) FILTER (WHERE le.sender = '@zigger:beeper.com') >= 3
       AND COUNT(DISTINCT le.sender) <= 2
),

-- Resolve display names via regex on BLOB content
-- (BLOBs use \x22 escapes instead of standard JSON quotes)
beeper_names AS (
    SELECT DISTINCT ON (le.room_id)
        le.room_id,
        regexp_extract(
            CAST(le.content AS VARCHAR),
            'displayname\\x22:\\x22([^\\]+)',
            1
        ) AS display_name
    FROM beeper.local_events le
    WHERE le.type = 'm.room.member'
      AND le.sender <> '@zigger:beeper.com'
      AND le.state_key <> '@zigger:beeper.com'
      AND CAST(le.content AS VARCHAR) NOT LIKE '%bridge bot%'
    ORDER BY le.room_id, le.event_ts DESC
),

beeper_final AS (
    SELECT
        bd.room_id AS id,
        COALESCE(NULLIF(bn.display_name, ''), 'unnamed') AS contact_name,
        bd.my_msg_count, bd.network, 'beeper_db' AS source
    FROM beeper_dms bd
    LEFT JOIN beeper_names bn ON bd.room_id = bn.room_id
),

-- iMessage DMs (style != 43 excludes groups)
imessage_dms AS (
    SELECT
        c.chat_identifier AS id,
        COALESCE(NULLIF(c.display_name, ''), c.chat_identifier) AS contact_name,
        COUNT(*) AS my_msg_count,
        'iMessage' AS network,
        'imessage_db' AS source
    FROM imessage.chat c
    JOIN imessage.chat_message_join cmj ON c.ROWID = cmj.chat_id
    JOIN imessage.message m ON cmj.message_id = m.ROWID
    WHERE m.is_from_me = 1 AND c.style <> 43
    GROUP BY c.ROWID, c.chat_identifier, c.display_name
    HAVING COUNT(*) >= 3
)

SELECT * FROM beeper_final
UNION ALL
SELECT * FROM imessage_dms
ORDER BY my_msg_count DESC;

Persisted DM Landscape Table

Results are materialized in ~/i.duckdb as dm_landscape:

-- Quick summary
duckdb ~/i.duckdb -c "
SELECT network, COUNT(*) AS threads, SUM(my_msg_count) AS total_my_msgs
FROM dm_landscape GROUP BY ALL ORDER BY threads DESC;
"
-- Signal: 232 threads (8,569 msgs)
-- iMessage: 119 threads (9,192 msgs)
-- Telegram: 9 threads (156 msgs)
-- Total: 360 threads, 17,917 messages

Schema Reference

beeper.local_events (Matrix protocol):

room_id TEXT, event_id TEXT, sender TEXT, type TEXT,
state_key TEXT, content BLOB, redacts TEXT, unsigned BLOB,
stream_order INT, event_ts INT, created_ts INT

Key event types:

• m.room.message — chat messages (sender, content with body)
• m.room.member — join/leave (content has displayname, membership)
• m.room.name — room name changes

Content BLOB quirk: uses \x22 instead of " — use regexp_extract not json_extract_string.

imessage.chat:

ROWID, chat_identifier, display_name, style (43=group)

imessage.message:

ROWID, text, is_from_me, date, handle_id

Join via imessage.chat_message_join(chat_id, message_id).

Search Guidelines

MCP queries are LITERAL WORD MATCHING, not semantic search.

• RIGHT: query="dinner" or query="flight"
• WRONG: query="dinner plans tonight"

Multiple words = ALL must match. Use single keywords.

For semantic/complex searches, use Tier 3 (DuckDB full-text search on SQLite archive).

Tier Selection Decision Tree

Need to SEND something?
  └─ Text only → Tier 1 (MCP)
  └─ File/attachment → Tier 1 (Desktop API HTTP: upload + send; `scripts/beeper_send_file.sh`)

Need chat type (DM vs group)?
  └─ Yes → Tier 2 (beeper-cli has type: "single"|"group")

Need full history or cross-platform JOIN?
  └─ Yes → Tier 3 (SQLite→DuckDB)

Need real-time / recent?
  └─ Yes → Tier 1 (MCP) or Tier 2 (beeper-cli)

Need contact name resolution?
  └─ MCP has senderName in messages
  └─ beeper-cli has title field on chats
  └─ SQLite has m.room.member displayname (most complete)

Workflow Patterns

Find and Message Someone

1. mcp__beeper__search_chats query="person name" → get chatID
2. mcp__beeper__list_messages chatID="..." → verify identity
3. mcp__beeper__send_message chatID="..." text="..."

Send a File to a Chat

1. Get the chatID via mcp__beeper__search_chats.
2. Send the file via scripts/beeper_send_file.sh.

DM Landscape Analysis (Tier 3)

duckdb ~/i.duckdb -c "
SELECT contact_name, my_msg_count, network
FROM dm_landscape
ORDER BY my_msg_count DESC LIMIT 20;
"

Cross-Platform Contact Lookup (Tier 3)

-- Find someone across Signal + iMessage + Telegram
INSTALL sqlite; LOAD sqlite;
ATTACH '~/Library/Application Support/BeeperTexts/account.db' AS beeper (TYPE sqlite, READ_ONLY);
ATTACH '~/Library/Messages/chat.db' AS imessage (TYPE sqlite, READ_ONLY);

-- Search Beeper (Signal/Telegram)
SELECT DISTINCT
    regexp_extract(CAST(content AS VARCHAR),
        'displayname\\x22:\\x22([^\\]+)', 1) AS name,
    room_id
FROM beeper.local_events
WHERE type = 'm.room.member'
  AND CAST(content AS VARCHAR) LIKE '%SearchName%';

-- Search iMessage
SELECT chat_identifier, display_name
FROM imessage.chat
WHERE display_name LIKE '%SearchName%'
   OR chat_identifier LIKE '%SearchName%';

Filter by Network (Tier 2)

# Get account IDs
BEEPER_ACCESS_TOKEN=$(fnox get BEEPER_ACCESS_TOKEN --age-key-file ~/.age/key.txt) \
  beeper-cli accounts list -o json

# Filter chats to WhatsApp only
BEEPER_ACCESS_TOKEN=$(fnox get BEEPER_ACCESS_TOKEN --age-key-file ~/.age/key.txt) \
  beeper-cli chats list -o json --account-ids whatsapp

Refresh dm_landscape Table

duckdb ~/i.duckdb -c "
INSTALL sqlite; LOAD sqlite;
ATTACH '\$HOME/Library/Application Support/BeeperTexts/account.db' AS beeper (TYPE sqlite, READ_ONLY);
ATTACH '\$HOME/Library/Messages/chat.db' AS imessage (TYPE sqlite, READ_ONLY);

CREATE OR REPLACE TABLE dm_landscape AS
WITH
beeper_dms AS (
    SELECT le.room_id,
        COUNT(*) FILTER (WHERE le.sender = '@zigger:beeper.com') AS my_msg_count,
        COUNT(DISTINCT le.sender) AS sender_count,
        CASE WHEN le.room_id LIKE '%.local-signal.%' THEN 'Signal'
             WHEN le.room_id LIKE '%.local-telegram.%' THEN 'Telegram'
             ELSE 'Other' END AS network
    FROM beeper.local_events le WHERE le.type = 'm.room.message'
    GROUP BY le.room_id
    HAVING COUNT(*) FILTER (WHERE le.sender = '@zigger:beeper.com') >= 3
       AND COUNT(DISTINCT le.sender) <= 2
),
beeper_names AS (
    SELECT DISTINCT ON (le.room_id) le.room_id,
        regexp_extract(CAST(le.content AS VARCHAR),
            'displayname\\\\x22:\\\\x22([^\\\\]+)', 1) AS display_name
    FROM beeper.local_events le
    WHERE le.type = 'm.room.member'
      AND le.sender <> '@zigger:beeper.com' AND le.state_key <> '@zigger:beeper.com'
      AND CAST(le.content AS VARCHAR) NOT LIKE '%bridge bot%'
    ORDER BY le.room_id, le.event_ts DESC
),
beeper_final AS (
    SELECT bd.room_id AS id,
        COALESCE(NULLIF(bn.display_name, ''), 'unnamed') AS contact_name,
        bd.my_msg_count, bd.network, 'beeper_db' AS source
    FROM beeper_dms bd LEFT JOIN beeper_names bn ON bd.room_id = bn.room_id
),
imessage_dms AS (
    SELECT c.chat_identifier AS id,
        COALESCE(NULLIF(c.display_name, ''), c.chat_identifier) AS contact_name,
        COUNT(*) AS my_msg_count, 'iMessage' AS network, 'imessage_db' AS source
    FROM imessage.chat c
    JOIN imessage.chat_message_join cmj ON c.ROWID = cmj.chat_id
    JOIN imessage.message m ON cmj.message_id = m.ROWID
    WHERE m.is_from_me = 1 AND c.style <> 43
    GROUP BY c.ROWID, c.chat_identifier, c.display_name HAVING COUNT(*) >= 3
)
SELECT * FROM beeper_final UNION ALL SELECT * FROM imessage_dms;

SELECT network, COUNT(*) AS threads, SUM(my_msg_count) AS total_my_msgs
FROM dm_landscape GROUP BY ALL ORDER BY threads DESC;
"

Resource-Aware Processing

NEVER pull full message history into context. Tier hierarchy:

1. Check DuckDB first (Tier 3) — zero network cost
2. Use MCP for recent (Tier 1) — bounded by limit param
3. Use beeper-cli for metadata (Tier 2) — chat types, accounts

Context Budget

CONTEXT_BUDGET = 10000 chars
Always set limit param on MCP calls.
Always use dateAfter on search_message_contents.
Prefer DuckDB queries over fetching raw messages.

GF(3) Triadic Access Pattern

Conservation: all three tiers complement without overlap. Desktop API writes (MCP + HTTP), beeper-cli routes, DuckDB archives.

Known Issues

• beeper-cli pagination cycles: cursor-based pagination can return duplicates; always deduplicate on chat ID
• Beeper BLOB encoding: content column uses \x22 escapes — use regexp_extract, not json_extract_string
• iMessage chat.db SQL: use <> not != (shell escaping issue)
• WhatsApp in account.db: WhatsApp bridge stores data differently; not present in account.db local_events — use beeper-cli for WhatsApp

MCP Server Config

{
  "beeper": {
    "command": "/bin/sh",
    "args": [
      "-c",
      "BEEPER_ACCESS_TOKEN=$(fnox get BEEPER_ACCESS_TOKEN --age-key-file ~/.age/key.txt) exec npx -y @beeper/desktop-mcp"
    ]
  }
}

Requires: fnox, age key at ~/.age/key.txt, npx in PATH, Beeper Desktop running.