The Agent Skills Directory

[SAFE]: The skill consists of documentation for legitimate email management tools. No obfuscation, data exfiltration, or malicious commands were found during analysis.
[COMMAND_EXECUTION]: The skill documentation describes the use of gog and himalaya to perform email-related tasks such as reading, searching, and sending messages. These commands are consistent with the skill's intended use case.
[PROMPT_INJECTION]: The skill provides the ability to read untrusted data from emails, which introduces an indirect prompt injection attack surface.
Ingestion points: Untrusted email content is ingested via gog mail read and himalaya read as shown in SKILL.md.
Boundary markers: The skill does not define specific markers or instructions to separate email content from agent instructions.
Capability inventory: The skill allows for various email operations, including sending, replying, and folder management.
Sanitization: No evidence of content sanitization or validation was found in the skill's instructions.
[NO_CODE]: The skill contains no executable scripts or binaries, relying solely on markdown-based documentation for its functionality.

c-email