The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill installs the @membranehq/cli package from the public NPM registry. This is the official command-line tool provided by the vendor for managing integrations.
[COMMAND_EXECUTION]: The instructions utilize various CLI commands (membrane login, membrane connection, membrane action) to facilitate the integration. These commands handle authentication securely by initiating browser-based flows rather than requesting or storing raw credentials.
[INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of data from the KiotViet API. While processing external data presents a theoretical surface for indirect prompt injection, the skill mitigates this by using structured JSON output (--json) and official pre-built actions which reduce the risk of the agent misinterpreting data as instructions.
Ingestion points: Tool output from membrane action run and membrane request (SKILL.md).
Boundary markers: None explicitly defined in the prompt instructions, though the use of machine-readable JSON formats acts as a natural data delimiter.
Capability inventory: The skill uses network operations mediated through the Membrane CLI; it does not perform direct file-system writes or dynamic code evaluation.
Sanitization: The skill relies on the Membrane platform's internal handling of API responses.

kiotviet