service-itsm-channels-coordinate

Warn

Audited by Gen Agent Trust Hub on Aug 27, 2026

Risk Level: MEDIUMDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [DYNAMIC_EXECUTION]: The skill dynamically identifies and invokes sub-skills at runtime by querying the environment's skill catalog using regex patterns (e.g., ^service-itsm-slack-.*-configure$). This method of resolving and executing code based on identifiers computed at runtime constitutes dynamic loading.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes user input and external environmental data (the skill catalog) to drive its workflow, which creates a potential surface for indirect prompt injection if the catalog metadata or user strings influence the execution path.
  • Ingestion points: User input for menu selection (choices 1-5) and the results returned from the internal skills catalog search.
  • Boundary markers: The skill includes basic input validation, but does not define explicit boundary markers or isolation for content retrieved from the skills catalog.
  • Capability inventory: The skill has the capability to perform administrative configuration writes via the dispatch tool and can invoke any other installed skills.
  • Sanitization: Includes logic to verify that user input matches the expected numerical menu options.
  • [COMMAND_EXECUTION]: The skill performs direct administrative writes to the organization's configuration using the dispatch tool for the Notifications setup feature. It executes authenticated PATCH requests to enable various preference toggles (Notifications, EmailNotifications, InAppNotifications, SlackNotifications, and TeamsNotifications) via the /services/data/v66.0/setup/org/preferences/ endpoint.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Aug 27, 2026, 04:35 PM
Security Audit — agent-trust-hub — service-itsm-channels-coordinate