Die Oberfläche ist lokalisiert; Skill-Beschreibungen und -Dokumente werden per KI automatisch in Ihre Sprache übersetzt. Code, Befehle, Pfade und Fachbegriffe bleiben im Original, maßgeblich ist das Dokument des Autors.
Skills entdecken
Entwicklung · SKILL.md

ttl-daily-marketing-review-agent

ttl-daily-marketing-review-agent

Perform a daily review of Texas Total Loss marketing activity, competitor updates, and performance to suggest next actions.

Independent GitHub snapshot · Automated classification · Original documentation governs

Skill-Dokumentation

GitHub

name: ttl-daily-marketing-review-agent description: Perform a daily review of Texas Total Loss marketing activity, competitor updates, and performance to suggest next actions. trigger: > Use this skill once per day (or on demand) to summarize what happened, what worked, and what to do next for Texas Total Loss marketing.

Goal

Create a concise daily intelligence report summarizing activity, performance, and recommending next actions.

Instructions

  1. Gather context
    • Read recent docs/marketing-insights.md, docs/competitor-trends.md, and campaign files.
  2. Review performance data
    • Look in data/analytics/ and summarize high-level metrics.
  3. Synthesize a Daily Summary
    • What changed, what performed best, issues/anomalies.
  4. Recommend next actions
    • Propose 3–5 actions, clearly ranked by impact/effort.
    • Apply Decision Tier Framework (from context.md):
      • Auto-Determine: If a choice has high confidence/precedent, state the decision made.
      • Propose Narrowed List: If multiple viable paths exist, present a "Top 3" list with trade-offs.
    • For each action, note:
      • Which skill(s) should handle it.
      • Whether it needs human review or can be mostly automated.
  5. Write back to workspace
    • Update docs/marketing-insights.md and .antigravity/brain/plan.md.
  6. Output a concise report.
Gespeicherte Skills