The Agent Skills Directory

[SAFE]: The skill is designed to use the official Membrane CLI for all operations, which is a secure method for handling authentication and API communication without requiring hardcoded secrets.
[COMMAND_EXECUTION]: The skill instructions involve executing membrane CLI commands for action discovery, creation, and execution. This behavior is aligned with the skill's purpose as an integration tool.
[EXTERNAL_DOWNLOADS]: The skill requires the installation of the @membranehq/cli package from the public npm registry. This package belongs to the vendor's official scope.
[PROMPT_INJECTION]: Indirect Prompt Injection Surface Analysis:
Ingestion points: The agent ingests data from DocRaptor actions via the output field of membrane action run and action metadata from membrane action list as documented in SKILL.md.
Boundary markers: No specific delimiters are used to wrap the output from the Membrane CLI.
Capability inventory: The agent has the capability to execute shell commands via the membrane CLI.
Sanitization: No explicit sanitization or validation of external action outputs is implemented within the skill instructions.

docraptor