The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill recommends installing the @membranehq/cli package from npm. This is an official tool provided by the skill's author to facilitate communication with their integration platform.
[COMMAND_EXECUTION]: The instructions guide the agent to use several membrane CLI commands for logging in, connecting to services, and executing actions. These commands are necessary for the skill's primary function of interacting with the CrowdStrike API via the Membrane gateway.
[PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection (Category 8). It takes natural language "intents" and JSON "inputs" (which may be derived from external sources) and passes them to CLI commands like membrane action list and membrane action run. If these inputs contain malicious instructions, they could potentially influence the agent's behavior or the actions performed within the CrowdStrike environment. No explicit sanitization or boundary markers are defined in the skill instructions.
Ingestion points: --intent parameter in membrane action list and --input parameter in membrane action run (SKILL.md)
Boundary markers: Absent
Capability inventory: Subprocess execution via shell (membrane CLI) and ability to read/modify security data in CrowdStrike
Sanitization: Absent

crowdstrike