Security audit
Security checks for vulnerabilities and agentic risk
This is a coherent mobile analytics helper that stores scoped local app notes and uses declared analytics APIs when credentials are provided.
Before installing, be comfortable with local storage of app analytics context under ~/mobile-app-analytics/. Do not put API keys, JWT private keys, access tokens, PII, or sensitive user-level exports in those files; use scoped platform credentials and restrict BigQuery/API access appropriately.
This pattern attempts to override system instructions or ignore safety constraints. Without LLM analysis, manual review is recommended.
Enable debug mode to see events in real-time:
# iOS Simulator
adb shell setprop debug.firebase.analytics.app com.your.app
Hidden instructions were detected in comments or invisible text. These could contain malicious directives. Manual review is recommended.
integration: pending | done | declined
## Apps Tracked
<!-- List each app you're helping analyze -->
<!-- app: AppName | platforms: iOS, Android | stack: Firebase, Mixpanel -->
## Current Focus
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
# Get app analytics
curl -X GET "https://api.appstoreconnect.apple.com/v1/apps/{app_id}/analyticsReportRequests" \
-H "Authorization: Bearer {jwt_token}"
The skill encourages exporting Firebase Analytics data to BigQuery and provides user-level query examples using identifiers like user_pseudo_id, but it gives no warning about privacy, access control, retention, or regulatory requirements. In an analytics context this can lead users to expose or overuse pseudonymous behavioral data without appropriate governance, increasing risk of privacy violations or unauthorized analysis.
The activation prompt is framed broadly enough that the skill may attach itself to many future conversations about general app metrics or analytics, increasing the chance of unintended invocation. In context, this is not overtly malicious, but it can cause the agent to act on user data or analytics topics persistently without sufficiently specific, informed consent.
The skill instructs the agent to save user preferences and integration status to a persistent file without first warning the user that information will be stored. Even though the stored content appears operational rather than highly sensitive, silent persistence creates a privacy and consent problem and can surprise users who did not agree to long-term retention.
The skill directs creation of per-app files containing app name, platforms, analytics stack, tracked events, funnel stages, and KPI targets, but the skill description does not warn users that this business metadata will be retained persistently. In context, this data can reveal internal product strategy and performance goals, so undisclosed storage raises confidentiality and privacy concerns beyond normal conversational assistance.
No suspicious patterns detected.