Analytics Agent

Analysiert Trading, App, Freelance und Marketing-Metriken, erstellt regelmäßige Reports, erkennt Trends und empfiehlt datenbasierte Optimierungen.

Security Scan

The skill's analytics description matches its instructions, but the runtime instructions are vague and imply accessing internal dashboards, logs, and sending reports without declaring what credentials or endpoints will be used — a mismatch that could expose sensitive data.

Purpose & Capability

The stated purpose (analytics for Trading, App, Freelance, Marketing) aligns with the SKILL.md steps (collect metrics, create reports, detect trends). However the instructions reference an internal dashboard (192.168.178.80:6969), 'Base44 Stats' (unclear service), log analysis and Excel/Sheets exports — items that suggest network access and service credentials not declared in the skill metadata.

Instruction Scope

SKILL.md tells the agent to analyze logs, query a local dashboard IP, aggregate data, and 'An Thomas senden' (send to a person) but provides no details about which credentials, mail endpoint, or storage to use. The instructions grant broad, ambiguous discretion (read logs, access dashboard, export/send data) which could cause the agent to access or exfiltrate sensitive internal data.

Install Mechanism

Instruction-only skill with no install spec and no code files — minimal disk footprint and no packaged install risk.

Credentials

The skill declares no required environment variables or credentials, yet its runtime steps imply the need for access tokens/credentials for dashboards, app analytics, Sheets/Excel exports, and an email/sending mechanism. The lack of declared credentials is a mismatch and means the agent may attempt to use other agent/system credentials or ask the user interactively, which is risky.

Persistence & Privilege

always is false and there are no install hooks or config paths requested. The skill does not request persistent or elevated platform privileges in its metadata.

What to consider before installing

This skill appears to do what it says (analytics + reports) but its runtime instructions are vague and expect access to internal dashboards, logs, and to send reports without declaring how the agent will authenticate. Before installing: (1) Ask the author for source code or a homepage and clarification of what endpoints and accounts the skill will access. (2) Require the skill to declare the exact credentials/scopes it needs (read-only tokens for analytics, explicit SMTP/API target for sending reports). (3) If you test it, run it in a sandbox or network segment that contains only non-sensitive data and block access to corporate systems. (4) Clarify the recipient 'Thomas' (email address/system) so reports don't get sent to an unexpected destination. (5) Prefer providing specific, minimum-privilege read-only credentials rather than letting the agent use any available account. If these clarifications are not provided, treat the skill as high-risk for data exposure and avoid enabling it on production accounts.

<objective> Tracking und Analyse aller unserer Projekte: Trading Performance, App Metrics, Freelance Leads, Marketing ROI. Erstellt Reports, identifiziert Trends, empfiehlt Optimierungen. </objective> <principles> ## Analytics-Fokus

Metriken

• Trading: PnL, Win-Rate, Bias-Performance
• App: Downloads, User, Revenue
• Freelance: Leads, Conversions, Revenue
• Marketing: Views, Engagement, Cost

Reports

• Täglich: Trading Summary
• Wöchentlich: Full Overview
• Monatlich: Strategy Review

Tools

• Dashboard (192.168.178.80:6969)
• Logs analysieren
• Excel/Sheets Export </principles>

<process> <step> <action>Trading Analytics</action> <details> - Paper vs Live PnL - Win-Rate pro Asset - Bias Performance check - Pattern WR tracken </details> </step> <step> <action>App Analytics</action> <details> - Base44 Stats abfragen - User onboarded? - Payments? - Errors? </details> </step> <step> <action>Freelance Tracking</email> <details> - Email-Response Rate - Angebote gesendet vs gewonnen - Revenue </details> </step> <step> <action>Report erstellen</action> <details> - Daten aggregieren - Trends identifizieren - Empfehlungen schreiben - An Thomas senden </details> </step> </process> <triggers> - Täglich (Trading Summary) - Wöchentlich (Full Report) - Nach Major Events </triggers>

<success_criteria>

• Alle Metrics im Blick
• Trends früh erkennen
• Bessere Decisions durch Data </success_criteria>