The Agent Skills Directory

[Operational Safety]: The skill implements a safety framework for AWS resource management by explicitly blocking destructive commands like delete-db-cluster and reset-graph, redirecting these to manual change-control processes. All create and modify operations require explicit user confirmation, reducing the risk of accidental infrastructure changes.
[Security Best Practices]: Instructions consistently mandate secure configurations, including encryption at rest via AWS KMS, encryption in transit using TLS 1.2+, and IAM database authentication (SigV4). It provides specific guidance on scoping security groups to trusted sources rather than open access.
[Input Validation and Parameterization]: The skill includes an input_validator.py script to sanitize resource identifiers using regular expressions. It emphasizes using parameterized queries for Gremlin and openCypher rather than string interpolation, which is an effective measure against query injection.
[Artifact Production]: The skill uses the file_write tool to save structured deliverables, such as JSON reports and query plans, to dedicated artifact paths. The instructions focus on efficient and correct execution of these tasks for downstream validation.
[Dependency Management]: External tools and libraries, such as the Apache TinkerPop Gremlin console and standard Python packages (boto3, gremlinpython), are referenced for legitimate setup and connectivity. The skill provides clear installation instructions that follow standard developer workflows.

amazon-neptune