quality-maturity-assessment
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external data such as delivery team plans, test strategies, and quality metrics. It addresses this inherent attack surface by instructing the agent to perform an 'Input Audit' to identify missing or conflicting information and to strictly separate confirmed facts from inferences, which reduces susceptibility to malicious data payloads.
- [COMMAND_EXECUTION]: The package includes a local evaluation script (
evals/fixtures/check-domain-decisions.sh) that invokes a Python script to validate test outcomes. This script is limited to parsing textual tables and comparing them against a static oracle for quality assurance purposes, posing no risk to the host system. - [SAFE]: The instructions explicitly include safety-oriented constraints, such as mandatory use of least privilege, data masking for sensitive environments, and a prohibition against hallucinating metrics, execution records, or release approvals. No evidence of obfuscation, remote code execution, or credential exposure was found.
Audit Metadata