The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses AWS CLI commands in SKILL.md and the boto3 library in scripts/agent.py to manage AWS Config recorders, delivery channels, and compliance rules. These operations are core to the skill's functionality for cloud infrastructure auditing.
[PRIVILEGE_ESCALATION]: The instructions involve creating IAM service-linked roles and defining remediation roles with permissions for AWS Config, Systems Manager (SSM), and Lambda. These high-privilege configurations are required for the skill to perform automated detection and fixing of non-compliant resources across an AWS Organization.
[INDIRECT_PROMPT_INJECTION]: The skill represents a surface for indirect prompt injection because it ingests untrusted metadata (such as resource tags and configuration items) from the AWS environment to evaluate compliance.
Ingestion points: The Lambda function in SKILL.md reads resource tags from configurationItem, and scripts/agent.py fetches compliance details from the AWS Config service.
Boundary markers: No specific delimiters or warnings are used to separate untrusted resource metadata from the rule evaluation logic.
Capability inventory: The skill includes functions to modify AWS configuration and remediation settings via put_config_rule and put_remediation_configurations across SKILL.md and scripts/agent.py.
Sanitization: The skill evaluates resource data without explicit sanitization before generating compliance annotations or reports.

implementing-aws-config-rules-for-compliance