The Agent Skills Directory

[SAFE]: The skill performs its stated purpose of executing web searches using Perplexity models via OpenRouter. No malicious code, persistence mechanisms, or privilege escalation attempts were found across any of the scripts or documentation.
[CREDENTIALS_UNSAFE]: The skill requires an OpenRouter API key for operation. It correctly advises users to store this key in environment variables or a .env file, providing a setup script (scripts/setup_env.py) to facilitate this securely. There are no hardcoded credentials in the codebase.
[EXTERNAL_DOWNLOADS]: The skill depends on the litellm Python package. This is a well-known, established library for interacting with various LLM providers. The documentation provides clear instructions for installation via standard package managers.
[DATA_EXFILTRATION]: Network communication is limited to the official OpenRouter API endpoints required for the skill's primary function. No evidence was found of the skill accessing sensitive local files (such as SSH keys or cloud credentials) or transmitting data to unauthorized third-party servers.
[COMMAND_EXECUTION]: The skill provides command-line interfaces for searching and environment setup. These scripts use standard argument parsing and do not utilize dangerous functions like eval() or os.system() on unvalidated user input.

perplexity-search