The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill requires the installation of the @membranehq/cli package via npm. This is a vendor-owned resource used for managing the integration.
[COMMAND_EXECUTION]: The skill relies on executing shell commands using the membrane CLI to list, create, and run actions against the PagerDuty API.
[DYNAMIC_EXECUTION]: The skill utilizes membrane action create, which generates new integration logic on the Membrane platform based on natural language descriptions. While this is a core platform feature, it involves dynamic generation of executable actions.
[INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted data from PagerDuty (such as incident notes, descriptions, or user-provided summaries) and presents it to the agent.
Ingestion points: Data returned from PagerDuty via membrane action run and membrane action list commands (e.g., incident details, log entries).
Boundary markers: No specific delimiters or instructions are provided to the agent to treat external PagerDuty data as untrusted or to ignore embedded instructions.
Capability inventory: The agent has the capability to execute further shell commands, create new actions, and run existing actions based on the processed data.
Sanitization: There is no evidence of sanitization or filtering of the content retrieved from the PagerDuty API before it is processed by the agent.

pagerduty