The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the user to execute various shell commands for environment setup, database management, and debugging.
Evidence includes instructions for bin/rails pay:install:migrations, bin/rails runner, and usage of the stripe CLI tool (e.g., stripe listen, stripe trigger). These are standard development operations for the Pay gem and Stripe integration.
[INDIRECT_PROMPT_INJECTION]: The skill handles data from external sources, creating an attack surface for indirect prompt injection via Stripe API responses and webhook events.
Ingestion points: External data enters the system through webhook handlers (e.g., app/webhooks/charge_succeeded_handler.rb mentioned in workflows/setup-webhooks.md) and controller parameters (e.g., params[:payment_method_id] in workflows/create-subscription.md).
Boundary markers: The skill does not implement or mention explicit boundary markers or instructions to disregard embedded commands in external data.
Capability inventory: The skill demonstrates capabilities for database modification (ActiveRecord models) and outgoing network requests to the Stripe API.
Sanitization: The workflows rely on standard Rails and Pay gem abstractions, which do not include specific sanitization against LLM-targeted injection attacks in the provided snippets.

pay-gem