The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute sqlite3 commands directly on the host system to interact with the application database.
[DATA_EXFILTRATION]: The skill identifies and accesses a highly sensitive local database path (~/Library/Application Support/BeeperTexts/index.db). This file contains private chat logs, contact hints, and message history from multiple bridged services (iMessage, WhatsApp). Reading this data into the agent's context is a significant privacy risk and represents data exposure.
[PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection by reading untrusted message content from external databases.
Ingestion points: Message content is retrieved from the mx_room_messages_fts table in index.db via shell commands.
Boundary markers: No delimiters or safety instructions are provided to help the agent distinguish between its instructions and the message content it retrieves.
Capability inventory: The skill leverages sqlite3 for data retrieval as seen in the provided workflow and probes.
Sanitization: There is no evidence of sanitization or filtering of the retrieved message content before it is processed by the agent.

beeper