The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were detected. The skill uses official vendor tools and adheres to safe credential handling practices.
[EXTERNAL_DOWNLOADS]: Fetches and installs the @membranehq/cli package from the official NPM registry, which is a required tool for the integration as provided by the vendor.
[COMMAND_EXECUTION]: Utilizes the membrane CLI to perform legitimate operations such as authentication, action discovery, and API requests to the Blend platform.
[PROMPT_INJECTION]: Identifies a surface for indirect prompt injection when processing data from the Blend API:
Ingestion points: API response data retrieved through membrane action run and membrane request (SKILL.md)
Boundary markers: Absent; no delimiters or warnings to ignore embedded instructions
Capability inventory: Shell execution of the membrane CLI (SKILL.md)
Sanitization: Absent; no validation or escaping of API content mentioned

blend