The Agent Skills Directory

[Infrastructure-as-Code Automation]: The skill drafts Terraform and shell scripts to assist in provisioning cloud resources. It implements a 'Plan-Validate-Execute' pattern which ensures that all drafted code is validated for syntax and then submitted as a Pull Request for user review, rather than being applied directly to the environment.
[Workspace Interaction]: The skill includes functionality to scan the user's workspace for existing database resource scripts. To ensure user control, it is instructed to ask for explicit confirmation of target file paths and directories before attempting any modifications.
[Input Processing Surface]: The skill processes user requirements from conversation history to provide recommendations and draft infrastructure code. While this represents an ingestion point for untrusted data, the risk is addressed through structured recommendation matrices and mandatory review processes.
Ingestion points: Discovery requirements are gathered from the user conversation as specified in SKILL.md and references/onboarding_prompts.md.
Boundary markers: The discovery phase is scoped to specific architectural categories, though explicit string delimiters are not defined in the prompt instructions.
Capability inventory: The skill utilizes scripts/database_onboarding_skill.py for local template formatting and possesses capabilities to read/write files in the workspace and initiate Pull Requests.
Sanitization: Recommendations are mapped to an internal list of specific product enums, and generated code is subjected to validation steps and human review prior to implementation.

cloud-databases-onboarding