The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: Fetches the @membranehq/cli package from the NPM registry, which is the official tool provided by the vendor for this integration. This is a standard installation of a vendor-owned resource.
[COMMAND_EXECUTION]: Executes shell commands via the membrane CLI to manage user login, connection life cycles, and run data operations against the TOTVS API.
[DATA_EXFILTRATION]: Transmits data to external endpoints associated with TOTVS and Membrane for record management. The skill encourages best practices by directing users to use platform-managed authentication rather than storing secrets locally.
[PROMPT_INJECTION]: The skill processes untrusted data from TOTVS records, which creates an attack surface for indirect prompt injection.
Ingestion points: TOTVS record data retrieved via membrane action run (SKILL.md).
Boundary markers: Absent; there are no specific instructions for the agent to use delimiters or ignore instructions within the retrieved data.
Capability inventory: The agent has the ability to execute various shell commands through the membrane CLI, including creating and running new actions.
Sanitization: No explicit sanitization, validation, or filtering of the external record content is described in the skill instructions.

totvs