use-railway

Pass

Audited by Gen Agent Trust Hub on Nov 18, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill makes extensive use of the railway CLI and executes complex queries via SSH on remote database instances (PostgreSQL, MySQL, Redis, and MongoDB) to gather performance metrics and perform maintenance tasks.
  • [EXTERNAL_DOWNLOADS]: The skill includes instructions to download and install the Railway CLI using shell scripts from the vendor's official domains (railway.com and agents.railway.com). These are standard installation methods for the tool.
  • [INDIRECT_PROMPT_INJECTION]: The database analysis scripts (analyze-postgres.py, etc.) fetch up to 1000 lines of unfiltered database logs into the agent's context.
  • Ingestion points: Database logs are fetched via the get_recent_logs function in dal.py and processed in the database-specific analysis scripts.
  • Boundary markers: The instructions do not specify the use of clear delimiters or 'ignore embedded instructions' warnings when presenting logs to the AI.
  • Capability inventory: The skill possesses high-privilege capabilities, including the ability to create/delete services, modify environment variables, and change database configurations.
  • Sanitization: Raw logs are ingested without filtering or sanitization. This surface area is an inherent part of the skill's diagnostic purpose, but creates a theoretical path for indirect injection if an attacker can write malicious instructions to the database logs.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 18, 2026, 04:30 PM