The Agent Skills Directory

[DATA_EXFLLTRATION]: The skill instructions involve the creation and reading of sensitive credential files and private keys.
Evidence:
Accesses and modifies the OCI configuration file located at ~/.oci/config.
Generates and manages a private RSA key at ~/.oci/oci_api_key.pem using openssl.
These operations are essential for the skill's primary purpose of OCI security management but involve high-value target files.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted user input and interpolates it into sensitive IAM policy statements and shell commands without explicit boundary markers or sanitization.
Ingestion points: SKILL.md uses placeholders for group names, compartment names, and OCIDs in policy templates and CLI command examples.
Boundary markers: No explicit delimiters or instructions are provided to the agent to ignore potentially malicious content within these user-supplied strings.
Capability inventory: The skill uses the oci Python SDK (identity.create_policy, identity.list_policies) and the oci CLI, which possess significant cloud infrastructure management permissions.
Sanitization: No evidence of input validation or escaping for the interpolated strings before they are sent to the Oracle Cloud API or executed via shell.

oraclecloud-security-basics