The Agent Skills Directory

[SAFE]: The skill files consist of instructional markdown and evaluation data without any executable scripts or dangerous code.- [PROMPT_INJECTION]: No patterns associated with prompt injection, jailbreaking, or system prompt extraction were found in the instructions or metadata.- [DATA_EXPOSURE]: There is no evidence of hardcoded credentials or access to sensitive system paths. The skill explicitly instructs users to avoid sensitive data in examples.- [EXTERNAL_DOWNLOADS]: All external references point to reputable domains (theleanstartup.com, leanstartup.co) associated with the skill's subject matter.- [COMMAND_EXECUTION]: No shell commands, privilege escalation attempts, or persistence mechanisms were detected.- [REMOTE_CODE_EXECUTION]: The skill does not attempt to download or execute external scripts or packages.- [OBFUSCATION]: The content is fully transparent and does not use Base64, zero-width characters, or other encoding techniques to hide information.

lean-startup