The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill accesses ~/.oci/config to initialize OCI SDK clients and authenticate against Oracle Cloud APIs. This file contains highly sensitive information, including user and tenancy identifiers and private authentication keys.
[EXTERNAL_DOWNLOADS]: Instructions specify upgrading packages oci, oci-cli, and certifi via the Python package manager. It also facilitates provider upgrades through terraform init, which downloads binary providers from the official Terraform Registry.
[COMMAND_EXECUTION]: The workflow relies heavily on executing shell commands such as pip, oci, terraform, and git. It includes bash scripts for scanning infrastructure code for deprecated resources and performing automated validation tests.
[DYNAMIC_EXECUTION]: Employs python3 -c to execute Python validation scripts and formatting logic directly from the command line, specifically for checking SDK status and parsing JSON metadata.
[INDIRECT_PROMPT_INJECTION]: The skill processes data from environment logs and package metadata, which could potentially be manipulated by external inputs.
Ingestion points: Reads and parses upgrade-plan.log, .terraform.lock.hcl, and requirements-oci-backup.txt.
Boundary markers: None; the skill does not use specific markers to delimit untrusted log data from agent instructions.
Capability inventory: The agent is granted capabilities to perform package installations, infrastructure changes via Terraform, and Git branch management.
Sanitization: None; the skill uses grep to filter for specific strings in log files without sanitizing the data to prevent potential instruction injection.

oraclecloud-upgrade-migration