The Agent Skills Directory

[COMMAND_EXECUTION]: The skill invokes python3 to execute a local audit log script (audit_log.py). It passes data derived from external sources, such as email subjects and recipients, as command-line arguments. This presents a risk of command injection if the input is not properly sanitized before being executed in a shell environment.
[DATA_EXFILTRATION]: The skill accesses sensitive local files, including ~/executive-assistant-skills/config/user.json, which contains email addresses, scheduling contacts, and signatures. While this is part of its functionality as an executive assistant, the access to local configuration files is a sensitive operation.
[PROMPT_INJECTION]: The skill contains an indirect prompt injection surface as it processes untrusted data from external sources.
Ingestion points: Reads email thread content from Gmail and meeting transcripts via the mcporter tool.
Boundary markers: Absent. The skill does not define delimiters or specific instructions to ignore embedded commands within the emails or transcripts it processes.
Capability inventory: The skill can create Gmail drafts, complete Todoist tasks, and execute local Python scripts via subprocess calls.
Sanitization: Absent. There is no instruction to sanitize or validate the content retrieved from external sources before using it to generate drafts or passing it to the audit logging script.

email-drafting