The Agent Skills Directory

[Indirect Prompt Injection] (HIGH): The skill is designed to process external user requests regarding Route53 configuration and possesses the capability to execute side-effecting commands on cloud infrastructure. This creates an exploitation surface where malicious input could trick the agent into deleting records, creating unauthorized entries, or exfiltrating AWS environment metadata.
Ingestion points: User-provided strings describing Route53 records and AWS patterns (SKILL.md).
Boundary markers: None identified in the provided documentation to isolate user input from command construction.
Capability inventory: Bash(aws:*) provides extensive write/execute access to the AWS environment.
Sanitization: No evidence of input validation or command escaping is provided in the skill manifest.
[Command Execution] (HIGH): The allowed-tools section explicitly permits Bash(aws:*). Providing an AI agent with the ability to execute shell commands targeting cloud infrastructure is a high-risk configuration, as any reasoning failure or prompt injection can result in immediate, non-trivial impact on the user's AWS account.

route53-record-manager