The Agent Skills Directory

[SAFE]: The skill consists entirely of Markdown documentation and templates. There are no executable scripts (Python, Node.js, Shell) or binaries included in the package.
[SAFE]: No network activity, remote script downloads, or patterns of remote code execution were found within the instructions or metadata.
[SAFE]: No sensitive file paths, credential harvesting patterns, or persistence mechanisms (like shell profile modification) were detected.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external email text, which is an inherent attack surface for indirect prompt injection.
Ingestion points: User-supplied email content is processed using the templates in templates/prompt-main.md.
Boundary markers: The prompts use descriptive placeholders like 'เนื้อหา original:' to separate data from instructions.
Capability inventory: The skill has no capabilities to execute code, make network requests, or write to the file system beyond creating the specified Markdown output.
Sanitization: No specific sanitization or filtering logic for external content is provided, but the lack of dangerous capabilities mitigates the risk.

email-manager