The Agent Skills Directory

[SAFE]: The skill provides a robust framework for competitive intelligence using standard documentation and configuration templates. It explicitly mandates adherence to ethical guidelines, such as using only public observable sources and avoiding paywall bypasses.
[COMMAND_EXECUTION]: The skill includes a local utility script scripts/verify.sh intended to be run by the agent to validate generated CSV and Markdown artifacts. The script is a standard bash linter that uses common utilities (grep, sed, find) to check for required data fields and sourcing without performing any network operations or privilege escalation.
[EXTERNAL_DOWNLOADS]: The skill references several well-known competitive intelligence tools and services (e.g., changedetection.io, Visualping) and provides a sample Docker Compose configuration. These references are for documentation and setup guidance, pointing to legitimate open-source repositories and established SaaS providers.
[PROMPT_INJECTION]: No malicious prompt injection patterns or attempts to bypass safety filters were detected. The instructions are task-oriented and include explicit 'ethics gate' checks to prevent misuse.

competitor-watch