The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires installing the @membranehq/cli package from the global NPM registry to facilitate communication with the Membrane platform.
[COMMAND_EXECUTION]: The instructions direct the agent to execute various shell commands using the membrane CLI for authentication (membrane login), connection management (membrane connect), and workflow execution (membrane action run).
[DATA_EXFILTRATION]: The skill retrieves structured data from external Make scenarios and modules via the membrane action run command, importing that data into the agent's execution context.
[PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it processes data returned from external Make actions which may be attacker-controlled.
Ingestion points: Data enters the context through the output of the membrane action run command as described in SKILL.md.
Boundary markers: There are no explicit delimiters or instructions provided to the agent to disregard potential instructions embedded within the retrieved Make data.
Capability inventory: The agent has the ability to execute shell commands, install software, and create new actions via the Membrane CLI as defined in SKILL.md.
Sanitization: No validation or sanitization steps are documented for the data received from the Make integration before it is interpreted by the agent.

make