The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Railway CLI and specialized scripts to manage infrastructure and introspect databases.
Scripts such as analyze-postgres.py and analyze-mysql.py execute commands via SSH and local subprocesses to gather metrics and logs.
State-modifying operations (e.g., enabling statistics or managing extensions) are explicitly restricted to user-only execution or require interactive terminal confirmation.
[EXTERNAL_DOWNLOADS]: The skill provides instructions for installing the official Railway CLI.
Methods include executing a vendor-provided install script via curl -fsSL cli.new | bash, as well as standard package managers like npm and brew.
All download sources are official Railway vendor resources.
[PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection through its database log ingestion feature.
The analyze-*.py scripts fetch up to 1,000 lines of raw, unfiltered database logs (recent_logs) for performance and error analysis.
While adversarial content could potentially be present in logs, the skill implements clear roles and structured analysis patterns to mitigate unintended obedience.

use-railway