The Agent Skills Directory

[CREDENTIALS_SAFE]: The skill manages API keys for various AI providers (OpenRouter, OpenAI, Anthropic, Google, Perplexity). It correctly instructs the agent to store these keys in a .env file and explicitly warns never to echo the keys back to the user, which is a standard security practice for credential management.
[EXTERNAL_DOWNLOADS]: The skill installs the voyage-geo package via pip. This package is the primary tool for the skill's functionality and originates from the skill's vendor (onvoyage-ai), representing normal vendor-provided functionality.
[COMMAND_EXECUTION]: The skill uses bash to execute voyage-geo CLI commands. These commands are limited to the intended purpose of performing brand research, query generation, and analysis. The user provides input for brand names and URLs, which are then passed into the CLI flags.
[INDIRECT_PROMPT_INJECTION]: The skill processes analysis results from JSON files (summary.json, analysis.json) generated by the CLI tool. While this represents a data ingestion surface, the risk is low as the data is structured and generated by the vendor's own analysis engine.
[SAFE]: All external URLs provided for API key management (OpenRouter, OpenAI, Anthropic, Google, Perplexity) are official, well-known service domains. No obfuscation or suspicious redirection was detected.

geo-run