configuring-experiment-rollout
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [SAFE]: Analysis of the skill instructions and reference materials revealed no malicious behavior, obfuscation, or unauthorized access attempts. The content is purely instructional and follows vendor best practices.
- [NO_CODE]: The skill consists solely of Markdown documentation and instructional guidelines. No executable scripts, binaries, or active code files are included in the package.
- [INDIRECT_PROMPT_INJECTION]: The skill handles user-provided data for experiment configuration (e.g., rollout ratios). It mitigates potential injection or misconfiguration risks by implementing defensive instructional patterns.
- Ingestion points: User-defined strings for rollout percentages and variant ratios (e.g., "60/40" or "25%") described in SKILL.md.
- Boundary markers: The skill explicitly directs the agent to clarify ambiguous inputs using specific "disambiguation questions" to confirm user intent.
- Capability inventory: Interactions involve experiment lifecycle management, feature flag updates, and exposure criteria configuration via the agent's platform tools.
- Sanitization: The instructions require mandatory warnings regarding statistical bias and user experience impact before any configuration changes are applied.
- [EXTERNAL_DOWNLOADS]: The skill references official PostHog technical documentation. These links are informational, clearly visible in plain text, and originate from the trusted vendor domain (posthog.com).
Audit Metadata