The Agent Skills Directory

[UNVERIFIABLE_DEPENDENCIES_AND_REMOTE_CODE_EXECUTION]: The skill requires the installation of external binaries (unifi-pp-cli and unifi-pp-mcp) from the author's NPM repository (@mvanhorn/printing-press-library) and GitHub repository (github.com/mvanhorn/printing-press-library). These dependencies are used to perform the core network auditing tasks and are fetched via npx and go install.
[DATA_EXPOSURE_AND_EXFILTRATION]: The CLI tool included with the skill supports a --deliver webhook:<url> flag. This feature allows the output of network audits or queries to be POSTed to a user-specified URL. The skill documentation presents this as a legitimate feature for machine-to-machine reporting and integration.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from a local SQLite database and the UniFi Network API. It uses a recall mechanism that maps natural language queries to stored command patterns (playbooks). While this processes external data, the system is designed to run locally and uses structured JSON envelopes for command output.
[SAFE]: The skill manages authentication via the UNIFI_API_KEY environment variable. It explicitly instructs users to generate this key from their gateway UI and store it locally, following standard secret management practices. No hardcoded credentials were found in the skill content.

pp-unifi