Vs. the Raw Decision Folder

What this means for you

The honest competitor is the agent’s own decision folder on GitHub. It works for one decision a week. It collapses on a hundred. LORE wins because reading agent output is a search problem, not a file-listing problem.

The pitch

Agents already write a decision file per action. You can absolutely read those files in a code editor. LORE replaces that habit with a searchable, filterable view that adds outcome tracking, cross-organization comparison, and a way to link verdicts back to learnings — none of which a folder of JSON files offers.

Who it’s for

The technical lead who can technically open the decision log directory and is wondering whether a dedicated reader is worth the bookmark.

Proof points

• Eight filter dimensions versus zero in a folder view: date, agent, role, type, severity, confidence, outcome, reviewed status
• Outcome annotation persists in a database column the JSON file does not have — files describe what an agent decided, LORE records what a human concluded
• Cross-organization analytics roll up multiple decision folders; opening the files in GitHub means opening one repository at a time

quadrantChart
  title LORE vs. ways people already read agent output
  x-axis "Browse files" --> "Search and filter"
  y-axis "Decision-only" --> "Decision plus outcome"
  quadrant-1 "LORE"
  quadrant-2 "Annotated wiki page"
  quadrant-3 "GitHub blob view"
  quadrant-4 "jq on the JSON folder"
  "GitHub blob view": [0.15, 0.2]
  "jq on the folder": [0.55, 0.25]
  "Spreadsheet export": [0.45, 0.55]
  "Annotated wiki page": [0.2, 0.65]
  "LORE": [0.85, 0.9]