debug-operation
Pass
Audited by Gen Agent Trust Hub on Aug 25, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the Shell tool to execute usdOptimize (a locally built binary) and ripgrep (for searching source code). These operations are restricted to the local build environment and are consistent with the skill's debugging purpose.
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to ingest and analyze untrusted external data, specifically USD assets and generated log files, creating a potential attack surface.
- Ingestion points: The agent reads asset files through USD Python APIs and triages run logs (e.g., debug-run.log).
- Boundary markers: There are no instructions to wrap these inputs in delimiters or use specific system instructions to ignore embedded commands within the files being analyzed.
- Capability inventory: The agent utilizes Shell, Read, Grep, and Glob tools to process these files and execute follow-up diagnostics.
- Sanitization: The workflow does not include steps for sanitizing log content or validating asset metadata before the agent processes them.
Audit Metadata