The Agent Skills Directory

[DATA_EXFILTRATION]: Accesses the macOS iMessage database located at ~/Library/Messages/chat.db. This file contains sensitive, private communication history.
Evidence: The script tools/imessage_parser.py identifies and attempts to read from the local chat.db path using the sqlite3 module.
Evidence: SKILL.md provides instructions for the agent to execute this script to ingest message data.
[COMMAND_EXECUTION]: Utilizes the Bash tool to execute local Python scripts that parse sensitive user files.
Evidence: SKILL.md directs the agent to run python3 scripts for parsing WeChat, iMessage, SMS, and photo metadata.
[PROMPT_INJECTION]: Ingests untrusted external data (chat logs, social media exports) to generate AI prompts, creating an indirect prompt injection surface.
Ingestion points: tools/*.py scripts read external files into the agent's context.
Boundary markers: Absent; the system prompts do not use delimiters or instructions to ignore embedded commands in the source data.
Capability inventory: The skill has access to Bash, Write, Edit, and Read tools.
Sanitization: No sanitization or validation of the content of the logs is performed before processing.

create-ex