The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructs the user to install the @membranehq/cli package globally via NPM. This is a vendor-provided tool required for the skill's functionality.
[COMMAND_EXECUTION]: The skill uses the membrane CLI to perform all operations. This includes login, connection management, action discovery, and executing API requests. This is the intended design for interacting with the Membrane ecosystem.
[DATA_EXFILTRATION]: While the skill facilitates data movement between the agent and Salesforce, it uses a proxy mechanism where authentication is managed server-side. This prevents the exposure of sensitive API keys or tokens in the local environment.
[PROMPT_INJECTION]: As the skill ingests data from Salesforce (such as record fields and search results), it possesses an indirect prompt injection surface.
Ingestion points: Data enters the context via membrane action run and membrane request outputs (SKILL.md).
Boundary markers: None are explicitly defined in the provided instructions.
Capability inventory: The skill can execute actions that write or modify data in Salesforce (create-record, update-record, etc.) and send arbitrary requests.
Sanitization: No explicit sanitization or validation of the retrieved data is mentioned before processing.

salesforce