The Agent Skills Directory

[PROMPT_INJECTION]: The skill defines a strategic advisor persona based on Andy Grove's management principles. The instructions are structured to guide the agent's response style and knowledge application. No malicious bypass patterns, safety filter overrides, or system prompt extraction attempts were detected.
[DATA_EXFILTRATION]: No commands for accessing sensitive file paths (e.g., .ssh, .aws, .env) or instructions to transmit data to external servers were identified. The skill operates within a strictly conversational and analytical scope.
[REMOTE_CODE_EXECUTION]: There are no instances of remote script downloads or piped shell execution (e.g., curl|bash). The installation command provided in the documentation is for the vendor's own package management and follows standard practices.
[OBFUSCATION]: Analysis of the markdown and reference files revealed no hidden text, zero-width characters, homoglyphs, or encoded strings designed to conceal malicious intent. All content is human-readable and matches its stated purpose.
[DYNAMIC_CONTEXT_INJECTION]: The SKILL.md file does not contain any shell execution patterns during skill loading. No usage of backtick-enclosed commands with the exclamation mark prefix was found.
[COMMAND_EXECUTION]: No instructions were found that would cause the agent to execute arbitrary shell commands or perform unauthorized system modifications.

grove-perspective