The Agent Skills Directory

[DATA_EXFILTRATION]: The skill provides instructions and Python code that access the sensitive OCI configuration file at ~/.oci/config. This file typically contains critical authentication credentials, including User and Tenancy OCIDs, fingerprints, and paths to private API keys used to manage cloud resources.
[EXTERNAL_DOWNLOADS]: The skill requires the installation of the oci Python package via pip. This is the official SDK for Oracle Cloud Infrastructure and is necessary for the skill's functionality.
[INDIRECT_PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection as it ingests data from OCI API responses and uses it to perform subsequent infrastructure management operations.
Ingestion points: External data enters the agent context via identity.list_availability_domains and network.list_subnets calls in SKILL.md.
Boundary markers: There are no explicit delimiters or instructions provided to the agent to distinguish between its own logic and potentially malicious data returned from the cloud environment APIs.
Capability inventory: The skill has broad capabilities to create, modify, and delete networking resources (VCNs, gateways, route tables, and security rules) using the OCI SDK.
Sanitization: No validation or sanitization is performed on the data retrieved from the OCI environment before it is used in logic or displayed to the user.

oraclecloud-core-workflow-b