By name: daily-briefing description: > This skill should be used when the user asks for a "daily briefing", "morning briefing", "news summary", "what's happening today", "daily digest", or wants a curated overview of news and trends across technology, business, policy, science, and culture. Also triggers when the user mentions "briefing plugin" or wants to understand the editorial philosophy behind the daily briefing system. version: 0.1.0
Daily Briefing System
A multi-agent pipeline that produces a curated daily briefing blending hard news with intellectually curious cross-domain content.
Editorial Philosophy
The briefing serves two audiences living in the same person:
The Pragmatist — wants to know what happened, what changed, and what it means for their work and decisions. Morning Brew's directness: no filler, sharp takes, efficient reading.
The Curious Intellectual — wants to understand why things happen and how seemingly unrelated domains connect. MIT Technology Review's depth: backstories, hidden connections, the "I never thought about it that way" moments.
Every briefing must serve both. Headlines and feature stories serve the pragmatist. The deep dive and curiosity corner serve the intellectual. The best pieces serve both simultaneously.
Pipeline Architecture
The briefing is produced by a 4-stage multi-agent pipeline (English + Turkish):
Stage 1: Domain Scouts (parallel)
Five specialized agents search their domains simultaneously:
- tech-scout — Technology, AI, software, hardware, open source
- business-scout — Markets, earnings, M&A, startups, economic indicators
- policy-scout — Regulation, trade, elections, geopolitics, governance
- science-scout — Research, health, biotech, climate, space
- culture-scout — Consumer trends, cultural phenomena, entertainment, social shifts
Each scout returns a structured list of 5-8 candidate stories ranked by importance and novelty.
Stage 2: Cross-Pollinator (editorial intelligence)
A single agent receives all scout outputs and performs editorial curation:
- Selects the best 8-12 stories across all domains
- Identifies 2-3 cross-domain connections
- Picks one story for deep-dive treatment
- Flags 2-3 curiosity pieces
- Outputs a structured editorial plan
Stage 3: Briefing Composer
A single agent takes the editorial plan and writes the final markdown briefing with consistent voice and structure.
Stage 4: Turkish Translation
After the English briefing is composed, it is translated to Turkish for daily.cantoramann.com/tr. A translation agent converts the full briefing while preserving its voice, rhythm, and personality — writing as a native Turkish speaker would, not as a textbook translation. Technical terms stay in English. The output is saved as briefing-YYYY-MM-DD-tr.md.
Briefing Sections
See references/briefing-structure.md for the full section-by-section specification.
Scout Strategies
See references/scout-strategies.md for domain-specific search strategies and quality criteria.
Editorial Standards
See references/editorial-philosophy.md for the full editorial guidelines, the curiosity content framework, and quality benchmarks.