diagnosing-failed-warehouse-syncs

Pass

Audited by Gen Agent Trust Hub on Apr 27, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill incorporates tools with the capacity to modify the state of data warehouse integrations, including destructive actions like external-data-schemas-resync (which wipes synced data) and external-data-schemas-delete-data. These operations are guarded by explicit instructions in Step 5 to obtain human confirmation before execution.
  • [DATA_EXPOSURE]: The workflow facilitates the management of sensitive authentication data, such as API keys and SSH configuration, via the external-data-sources-partial-update tool. This is the intended primary purpose of the skill (credential rotation and recovery) and aligns with standard administrative practices for the vendor (PostHog).
  • [PROMPT_INJECTION]: The skill analyzes latest_error strings retrieved from external data sources to classify failures. This constitutes an indirect prompt injection surface as error messages from external services could theoretically contain adversarial content.
  • Ingestion points: Error strings from external-data-sources-retrieve and external-data-schemas-list.
  • Boundary markers: None present for the processing of external error strings.
  • Capability inventory: The skill has access to tools for credential updates and destructive data operations across the data warehouse environment.
  • Sanitization: No explicit sanitization or escaping is performed; the skill relies on substring matching to identify known error patterns.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 27, 2026, 06:45 PM