The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the node -e command to perform read and write operations on a local file named health-data.json. These operations are used to store and update daily health metrics.
[INDIRECT_PROMPT_INJECTION]: The skill interpolates user-provided input (e.g., water intake amount) into shell command strings via the CUPS and NEW_CUPS placeholders. This creates a surface where a malicious user could attempt to inject code into the shell environment if the agent does not strictly validate the input as a number.
Ingestion points: User input for cup counts is passed into Node.js one-liners in SKILL.md.
Boundary markers: Absent; placeholders are placed directly within the command string without delimiters.
Capability inventory: The skill has file system access (read/write) via the Node.js fs module.
Sanitization: Absent; the instructions rely on the agent to correctly identify and provide a number rather than implementing code-level escaping or validation.

healthcheck