The Agent Skills Directory

[PROMPT_INJECTION]: No patterns of prompt injection, bypass instructions, or malicious overrides were detected. The skill uses instructions to guide a structured planning process without attempting to subvert agent safety guidelines.- [DATA_EXFILTRATION]: No evidence of credential harvesting or data exfiltration. The skill uses standard development tools like git and the GitHub CLI to interact with local repositories and authorized remotes, which is consistent with its stated purpose as an engineering project planner.- [REMOTE_CODE_EXECUTION]: The skill does not perform any remote code downloads or execution. It specifically claims to be a 'pure Markdown skill' with no executable scripts, hooks, or build steps. The installation instructions provide safe commands for manual source verification (git fetch/log/checkout).- [COMMAND_EXECUTION]: The skill references common CLI tools (git, gh, test) that are expected for its functionality. There are no instances of arbitrary command execution or suspicious shell usage.- [EXTERNAL_DOWNLOADS]: No suspicious external downloads or references to untrusted third-party services. The project identifies its origin as 'community' and refers to a reference design on GitHub without executing content from that source.

blueprint