performance-regression-analysis
Pass
Audited by Gen Agent Trust Hub on Oct 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external data such as performance metrics, logs, traces, and JSON/YAML result sets. While this data is untrusted, the skill instructions (in
prompts/performance-regression-analysis.md) mandate a strict 'Input Audit' and 'evidence basis' for every conclusion, which serves as a defensive boundary against interpreting embedded instructions as facts. - Ingestion points: Processes user-supplied baseline/candidate results and workload configurations.
- Boundary markers: Instructions explicitly require separating confirmed facts from working assumptions.
- Capability inventory: The skill primarily performs data analysis and generates Markdown/JSON artifacts; no unauthorized network or file-write capabilities are requested.
- Sanitization: The 'Quality Bar' section enforces that the agent must never invent system behavior or numbers.
Audit Metadata