The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes external text data for translation and rephrasing, which introduces a surface for indirect prompt injection where instructions hidden in the processed text could influence the agent.
Ingestion points: Untrusted data enters the context via input parameters for actions like translate-text and rephrase-text defined in SKILL.md.
Boundary markers: Absent; the instructions do not include delimiters or warnings to isolate input text from agent instructions.
Capability inventory: The skill uses CLI commands (membrane action run) and proxy requests (membrane request) to perform network operations.
Sanitization: Absent; no sanitization, filtering, or validation of the input text is described in the skill content.
[COMMAND_EXECUTION]: Instructs the user to execute various terminal commands using the membrane CLI utility to perform authentication, discover API actions, and run machine translation tasks.
[EXTERNAL_DOWNLOADS]: Recommends the global installation of the @membranehq/cli package from the NPM registry, which is the official tool provided by the vendor for managing integrations.

deepl