How it works
Items remain visible with their evidence and terms while repeated themes are reviewed and promoted into a clearly named topic when warranted.
- Capture item
- Review terms
- Detect pattern
- Promote when stable
Key concepts
- Emerging theme
- A repeated signal without an established category.
- Taxonomy review
- Periodic checking for categories that should split or merge.
- Promotion
- Moving a mature theme into its own durable topic.
Engineering risks
- A broad bucket can conceal important patterns.
- Premature categories make navigation noisy.
- Unreviewed items can remain uncategorised indefinitely.
Aggregated from the site
Evidence and developments
Matching public material uses the same topic rules as AI Engineering.
public GitHub discoveries
lostlight530/china-agentic-observatory
public GitHub discoveries
lostlight530/agentic-frontier-observatory
public GitHub discoveries
MaxonT/awesome-agent-workflows
public GitHub discoveries
git4rajmohan/AI_Projects
public GitHub discoveries
jaccen/Awesome-Agent-Engineering
public GitHub discoveries
ybak/ai-tutorial
public GitHub discoveries
kennethlaw325/awesome-llm-knowledge-systems
public GitHub discoveries
harvard-edge/cs249r_book
public GitHub discoveries
NikolaiSachok/ai-engineering-handbook
public GitHub discoveries
KaiWU5/Awesome-AI4AI
public GitHub discoveries
kartikm7/kratos
public GitHub discoveries
snowbitx/agenthub
public GitHub discoveries
donnyboy72/Junior-Dev-to-Pro
public GitHub discoveries
sungeuns/foundation-model-engineering
public GitHub discoveries
js-lee-AI/awesome-llm-agent-papers
public GitHub discoveries
MrPeppersDev/agent-infrastructure-landscape
public GitHub discoveries