The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides instructions to download a binary from the author's GitHub repository at https://github.com/dl-alexandre/UniFi-Site-Manager-CLI/releases/latest/download/usm-darwin-arm64.
[REMOTE_CODE_EXECUTION]: The manual installation steps involve downloading a remote binary, making it executable, and moving it to a system path (curl -L [URL] -o usm; chmod +x usm; sudo mv usm /usr/local/bin/).
[COMMAND_EXECUTION]: The skill instructions require the use of elevated privileges (sudo) to install the downloaded binary into a protected system directory.
[PROMPT_INJECTION]: The skill identifies a surface for indirect prompt injection by processing external data from a remote API.
Ingestion points: Data enters the agent context from the UniFi Site Manager API (api.ui.com) via the usm sites list and usm sites get commands as documented in SKILL.md.
Boundary markers: There are no explicit delimiters or instructions provided to the agent to ignore or isolate potential commands embedded within the API responses.
Capability inventory: The agent executes the usm CLI tool which interacts with both the local filesystem (configuration) and network APIs.
Sanitization: There is no evidence of sanitization or validation of the data fetched from the external API before it is presented to or processed by the agent.

usm-cli