qe-test-execution
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data, creating a surface for indirect prompt injection attacks.
- Ingestion points: The skill processes
gitChanges(mentioned inSKILL.md) and analyzes pull request details (referenced inevals/qe-test-execution.yaml), which are externally controlled inputs. - Boundary markers: There are no explicit delimiters or specific instructions (e.g., 'treat the following data as text only') to prevent the agent from following instructions that might be embedded within the code changes or test reports being analyzed.
- Capability inventory: The skill has the capability to execute shell commands using the
aqeCLI and orchestrate workflows within CI/CD environments like GitHub Actions. - Sanitization: There is no evidence of input sanitization or escaping for the natural language data interpolated into the agent's tasks.
- [EXTERNAL_DOWNLOADS]: The skill utilizes external resources and tools common in development environments.
- The
evals/qe-test-execution.yamlandSKILL.mdfiles reference official GitHub Actions (actions/checkout@v4,actions/upload-artifact@v4) and standard command-line utilities likejq. These are well-known and trusted services in the context of CI/CD orchestration.
Audit Metadata