Architecture

End-to-End Pipeline

flowchart LR
  A["Phase 1A URL Manifest"] --> B["Phase 1B Content Retrieval"]
  B --> C["Phase 1C Consolidation"]
  C --> N["Phase 1.5 Curator Notes (optional)"]
  N --> D["Phase 2 Event Sources"]
  D --> E["Phase 2 Events"]
  E --> F["Phase 3 Content Curation"]
  F --> G["Phase 4 Assembly"]
  G --> H["Phase 4.5 Polishing"]
  H --> I["Validation + Scoring"]

Self-Learning Propagation

flowchart LR
  A["Human corrections"] --> B["Learning capture"]
  B --> C["Skill + reference updates"]
  C --> D["Regenerate"]
  D --> E["Validate + score"]
  E --> F["Future runs improve"]

Interfaces and Artifacts

• Agent entrypoint: .github/agents/customer_newsletter.agent.md
• Phase skills: .github/skills/*/SKILL.md
• Phase prompts: .github/prompts/*.prompt.md
• Fresh-cycle prep: tools/prepare_newsletter_cycle.sh
• Strict validation: tools/validate_pipeline_strict.sh
• Deterministic event sources:
• kb/EVENT_SOURCES.yaml
• tools/extract_event_sources.py
• Output sample: output/2026-02_february_newsletter.md