qemu-tcg-frontend
Pass
Audited by Gen Agent Trust Hub on Jul 31, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or unauthorized network activity detected. The instructions follow professional development practices focused on technical correctness.
- [COMMAND_EXECUTION]: The skill performs local build and test commands (e.g.,
tests/tcg/) to verify instruction translations. It also uses git commands to manage repository-local exclusion files to ensure agent-specific metadata is not committed to the source repository. - [DATA_EXFILTRATION]: Workspace metadata and command results are recorded locally in a hidden folder (
.oh-my-qemu/). The skill policy requires redacting these logs to prevent unintended data exposure. - [PROMPT_INJECTION]: The skill handles untrusted instruction specifications and workspace source files.
- Ingestion points: Guest ISA definitions and target source code in
target/<arch>/. - Boundary markers: Structured logging in
audit.mdand clear documentation requirements to separate observations from inferences. - Capability inventory: File system writes, QEMU build processes, and test binary execution.
- Sanitization: Explicit instructions to redact command strings and results in
commands.md.
Audit Metadata