ai-engineering
Pass
Audited by Gen Agent Trust Hub on May 17, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: Several files mention prompt injection patterns (e.g., 'Ignore previous instructions'), but these are documented as security risks to mitigate or reject, rather than instructions for the agent to follow. This is a defensive security feature of the skill.
- [COMMAND_EXECUTION]: The skill uses the
Skill()function to compose functionality with other specialized skills (e.g.,otel-instrumentation). This is a standard platform orchestration pattern for modular agents and is used here for observability purposes. - [EXTERNAL_DOWNLOADS]: The skill references a wide range of well-known industry packages and services for testing, observability, and AI development (e.g., PyPI, NPM, OpenTelemetry, Dash0). These references are for engineering guidance and do not represent malicious downloads.
- [DATA_EXFILTRATION]: The skill provides instructions on PII redaction and least-privilege tool design to prevent data exfiltration in AI applications. No logic for unauthorized data collection or external transmission was found.
- [SAFE]: The skill's primary purpose is advisory. It guides users through complex AI engineering tasks using curated industry research and proven patterns from established providers like Anthropic and OpenAI.
Audit Metadata