The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the installation of the @membranehq/cli package from the official NPM registry. This is a vendor-provided tool used to interact with the Membrane platform.- [COMMAND_EXECUTION]: The agent is instructed to use various subcommands of the membrane CLI, such as login, connection, and action, to manage the integration. This involves executing shell commands to perform platform operations.- [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface (Category 8) because it processes data from external actions. \n
Ingestion points: Data enters the context through the output of membrane action list and membrane action run commands as described in SKILL.md. \n
Boundary markers: The instructions do not specify any markers or delimiters to isolate data retrieved from external actions from the agent's instructions. \n
Capability inventory: The skill allows the agent to execute actions on external systems via the membrane action run command. \n
Sanitization: There is no mention of sanitizing or validating the data returned from the Membrane platform before it is used by the agent.
[METADATA_ANALYSIS]: The official documentation link provided (developer.apple.com/documentation/clockkit/...) refers to an Apple Watch feature rather than the Breathe HR platform described in the skill text. This appears to be a documentation error rather than a malicious deception.

breathe