The Agent Skills Directory

Command Execution: The skill instructs the agent to run bun run ci to verify the codebase after edits. This is a common development practice to ensure that changes do not negatively impact the application.
Modification of Application Logic: The skill guides the agent to remove authentication and trial-related restrictions, such as email domain filters and usage caps. While these are changes to the application's logic, they are the primary intended function of the skill for developers who want to self-host the software.
Indirect Prompt Injection Surface: The skill requires the agent to read and modify existing application source files, which introduces an entry point for data that could theoretically influence the agent's behavior.
Ingestion points: The agent reads multiple application files, including apps/web/app/api/sessions/route.ts and apps/web/lib/managed-template-trial.ts.
Boundary markers: The instructions do not specify any delimiters or ignore-markers for the content read from these files.
Capability inventory: The agent has permission to read/write files and execute the bun command.
Sanitization: The skill does not mention any sanitization or validation of the code content before it is processed by the agent.

remove-demo-limits