perf-analyzer
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process and synthesize external performance data, including baseline metrics, experiment results, and profiling evidence. This creates a potential surface where malicious instructions embedded within logs or performance data could influence the agent's output.
- Ingestion points: The 'Inputs' section of SKILL.md identifies baseline data, experiment results, profiling evidence, and breaking point results as data sources.
- Boundary markers: The skill does not provide explicit delimiters or instructions to the agent to distinguish between valid data and potential embedded instructions within the inputs.
- Capability inventory: The provided skill file contains instructions only and does not define any executable code, file system writes, or network operations, which limits the potential impact of an injection.
- Sanitization: There are no instructions for the agent to sanitize or validate the data ingested from external performance tools or logs.
Audit Metadata