The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the Membrane CLI (membrane) for all interactions with the SharpSpring API, including authentication, action discovery, and execution. This centralizes security controls within the Membrane platform and ensures that the agent does not handle raw credentials.\n- [EXTERNAL_DOWNLOADS]: Suggests the installation of the @membranehq/cli package from NPM. This is an official package provided by the skill author (membranedev) and is a standard requirement for this integration type. No downloads from untrusted third-party sources were detected.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted data from SharpSpring (e.g., lead details, account names). Ingestion points: SKILL.md (via data retrieval actions like list-leads and get-account). Boundary markers: Absent. Capability inventory: membrane action run and membrane request functionality described in SKILL.md. Sanitization: Not explicitly defined in the skill instructions. While this is a characteristic of most CRM integrations, users should be aware that data retrieved from the CRM could influence agent behavior if not properly handled by the underlying LLM.

sharpspring