oraclecloud-debug-bundle

Pass

Audited by Gen Agent Trust Hub on Sep 9, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the official OCI CLI to perform infrastructure diagnostic tasks. This includes capturing instance console history, managing instance-agent commands to retrieve remote logs, and performing metadata queries. These actions are aligned with the skill's stated purpose of cloud infrastructure debugging.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests external data from potentially untrusted sources, specifically log files and console output from remote OCI instances.
  • Ingestion points: Retrieves serial console logs and cloud-init output using CLI commands in SKILL.md (Steps 3 and 4).
  • Boundary markers: No specific delimiters or "ignore instructions" wrappers are applied to the ingested log data.
  • Capability inventory: The skill possesses capabilities for local file writing and Bash command execution.
  • Sanitization: Log content is stored directly into the debug bundle without sanitization or filtering.
  • [EXTERNAL_DOWNLOADS]: The skill requires the installation of the official oci Python package via pip. This dependency is a standard, well-known SDK provided by Oracle for interacting with its cloud services.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 9, 2026, 03:45 AM
Security Audit — agent-trust-hub — oraclecloud-debug-bundle