The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the @gethookmyapp/cli tool and several Node.js scripts (scripts/*.mjs) to perform account management and messaging tasks. These operations are essential to the skill's function and target the vendor's official API endpoints.
[EXTERNAL_DOWNLOADS]: The skill setup involves installing the @gethookmyapp/cli package from NPM and cloning a starter kit from the vendor's GitHub repository. Both actions involve trusted resources associated with the skill's author.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data such as incoming chat messages and Instagram comments, which presents a surface for indirect prompt injection.
Ingestion points: Untrusted data enters the agent's context through webhook listening (references/channels.md) and message retrieval commands (references/whatsapp.md, references/instagram.md).
Boundary markers: The instructions explicitly warn the agent in references/mcp.md to treat external content (like notification bodies) as data rather than instructions.
Capability inventory: The agent has the ability to execute CLI commands (SKILL.md), write to configuration files (SKILL.md), and perform network operations via bundled scripts (scripts/*.mjs).
Sanitization: The skill provides instructions for the agent to redact sensitive information and maintain clear boundaries between system instructions and user-provided data.

integrate-hookmyapp