rejseplanen
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill retrieves real-time transit data from the well-known Rejseplanen API domain at
https://www.rejseplanen.dk/api. - [INDIRECT_PROMPT_INJECTION]: The skill processes external information such as station names, trip details, and service alerts from the Rejseplanen API, which are then presented to the AI agent. This constitutes a standard surface for processing external content from a reputable source.
- Ingestion points:
cli/src/commands/trip.ts,cli/src/commands/departures.ts,cli/src/commands/arrivals.ts, andcli/src/commands/disruptions.ts. - Boundary markers: Data is structured as JSON, tables, or plain text, providing clear separation for the agent.
- Capability inventory: The skill performs network requests to official transit endpoints and executes local logic via the Bash tool.
- Sanitization: The CLI parses structured JSON responses and filters out internal technical notes before outputting data to the agent.
- [SAFE]: No evidence of prompt injection, data exfiltration, malicious obfuscation, or persistence mechanisms was found. Authentication via the
REJSEPLANEN_ACCESS_IDis correctly handled through environment variables, avoiding hardcoded secrets.
Audit Metadata