The Agent Skills Directory

[SAFE]: The skill is purely instructional and defines workflows for managing project-related Markdown files (e.g., todo.md, roadmap.md). No commands for data exfiltration, unauthorized file access, or remote code execution were found.
[COMMAND_EXECUTION]: The skill mentions executing standard development tools like lint-md and pnpm docs:check:i18n as part of the phase-end verification process. These are routine development operations and do not pose a security risk in this context.
[DATA_EXPOSURE_AND_EXFILTRATION]: File access is restricted to project documentation paths within the repository. No access to sensitive system files or credentials was detected.
[INDIRECT_PROMPT_INJECTION]: The skill processes data from project files (todo.md, roadmap.md). While this presents a theoretical surface for indirect prompt injection if those files contain untrusted input, the instructions focus on structured formatting and status synchronization, which minimizes the risk.

todo-manager