The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were detected in the skill instructions, scripts, or templates.
[COMMAND_EXECUTION]: The skill utilizes shell scripts and command hooks to automate status verification and file initialization. These operations are limited to project management tasks and do not involve unsafe user input or elevated privileges.
[INDIRECT_PROMPT_INJECTION]: The skill's architecture involves reading and processing markdown files created during the session. While these files are used as context, the logic for parsing them (e.g., checking task status) is handled safely and does not allow for command injection from the file content.
[DYNAMIC_CONTEXT_INJECTION]: The skill uses platform-native hooks to inject the current task plan into the agent's context before tool use, ensuring the agent remains focused on its goals without requiring the !command syntax.

planning-with-files