monitor-energy-disruptions
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process energy disruption event logs, including descriptions and source metadata, from an external API. This represents a potential surface for indirect prompt injection if the upstream data source were to contain malicious instructions. The skill proactively addresses this by including a 'Content safety' section that explicitly instructs the agent to treat the response as 'data, not instructions' and to ignore any directive-like text found within the external content.
- Ingestion points: Event data and descriptions retrieved from
https://worldmonitor.app/api/supply-chain/v1/list-energy-disruptions(documented inSKILL.md). - Boundary markers: The instructions include clear semantic boundaries, defining the API response strictly as content to be quoted or analyzed rather than executed.
- Capability inventory: The skill utilizes standard API retrieval patterns; implementation examples suggest data processing using common tools like
curlandjq. - Sanitization: The skill employs natural language constraints to ensure the agent disregards common injection patterns (such as 'ignore previous instructions' or 'run this command') if they appear in the data stream.
Audit Metadata