The Agent Skills Directory

[SAFE]: The skill outlines architectural patterns for implementing framework-agnostic observability. No malicious patterns, credential theft, or unauthorized persistence mechanisms were identified.- [EXTERNAL_DOWNLOADS]: The skill utilizes official SDKs and endpoints for well-known services such as Firebase, Google Analytics, Microsoft Clarity, PostHog, and Sentry. These are recognized as trusted industry-standard tools.- [DYNAMIC_EXECUTION]: The provided code snippets use dynamic import() for the Firebase SDK (services/analytics/adapters.firebase.web.ts) to ensure SSR compatibility and optimize performance. This is a standard and safe web development practice.- [INDIRECT_PROMPT_INJECTION]: The skill accepts external data through a props parameter in the track function (services/analytics/track.ts). Evidence review: 1. Ingestion point: props parameter. 2. Boundary markers: Consent check hasConsent() at the fan-out boundary. 3. Capability inventory: No dangerous command execution or file-system write capabilities are present. 4. Sanitization: The skill mandates 'PII-light' properties and scrubbing of PII in error reports. Given the lack of dangerous capabilities, this surface is evaluated as safe.

frontend-observability