The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the agent to download and install the healthmd tool using Homebrew (brew install CodyBontecou/tap/healthmd) and by cloning a GitHub repository (https://github.com/CodyBontecou/health-md.git). Both sources are managed by the skill author.
[COMMAND_EXECUTION]: The skill relys on executing shell commands via the healthmd CLI to interact with an iPhone, query health metrics, and export data to the local filesystem.
[INDIRECT_PROMPT_INJECTION]: The skill ingests health data (JSON/JSONL) from an external iPhone device into the agent's processing context. This creates a surface where maliciously crafted strings within health records (e.g., workout notes or custom labels) could attempt to influence the agent's behavior.
Ingestion points: Data enters via the healthmd export, healthmd extract, and healthmd status commands.
Boundary markers: No specific delimiters or "ignore instructions" markers are defined for the agent to use when interpreting the exported health data.
Capability inventory: The agent can execute the healthmd CLI and read/write files to the local system.
Sanitization: There is no evidence of sanitization performed on the content of the health records before they are presented to the agent.

healthmd-cli