integration-javascript_node
Pass
Audited by Gen Agent Trust Hub on Mar 27, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill facilitates the analysis of 10-15 project files to identify business-critical events for tracking. This creates a surface for indirect prompt injection where untrusted content in the codebase could influence the agent's integration plan. \n
- Ingestion points: Project source code files (references/basic-integration-1.0-begin.md). \n
- Capability inventory: The skill utilizes file read/write operations and PostHog MCP tools for dashboard creation. \n
- Boundary markers: The instructions do not define specific boundary markers for the ingested code content. \n
- Sanitization: No explicit sanitization logic is prescribed for data extracted from project files before it is used in code generation or dashboard configuration.\n- [EXTERNAL_DOWNLOADS]: The documentation references standard installation procedures for official PostHog SDKs (
posthog-nodeand@posthog/ai) from established package registries. These resources are provided by the verified vendor and align with the skill's intended purpose.
Audit Metadata