The Agent Skills Directory

[SAFE]: The skill serves as a development framework and documentation for creating AI agents. All provided content, including system prompts and examples, is educational and follows safe implementation patterns.
[COMMAND_EXECUTION]: The skill includes a shell script scripts/validate-agent.sh used for local validation of agent file structures. The script uses standard utilities (grep, sed, awk) to perform text-based format checks without executing any user-supplied content or external code.
[PROMPT_INJECTION]: While the skill provides templates for defining agent personas and system prompts, these are presented as developer guides. No injection patterns aimed at bypassing agent safety filters or hijacking the current session were found.

Agent Development