Expand arXiv paper corpus

data/summary.json

@@ -1,8 +1,8 @@
 {
-  "total": 21,
+  "total": 989,
   "by_collection": {
     "jobs": 6,
-    "papers": 6,
+    "papers": 974,
     "industry": 9
   },
   "by_status": {
@@ -11,8 +11,8 @@
       "unknown": 2
     },
     "papers": {
-      "skimmed": 5,
-      "queued": 1
+      "queued": 969,
+      "skimmed": 5
     },
     "industry": {
       "analyzed": 8,
@@ -22,44 +22,72 @@
   "top_topics": [
     [
       "agent-evaluation",
-      8
+      837
+    ],
+    [
+      "tool-use",
+      728
+    ],
+    [
+      "rag",
+      460
+    ],
+    [
+      "reasoning",
+      378
+    ],
+    [
+      "planning",
+      347
+    ],
+    [
+      "agent-safety",
+      345
+    ],
+    [
+      "memory",
+      296
+    ],
+    [
+      "computer-use",
+      257
+    ],
+    [
+      "multi-agent",
+      245
+    ],
+    [
+      "workflow-agent",
+      237
+    ],
+    [
+      "coding-agent",
+      205
+    ],
+    [
+      "world-model",
+      83
+    ],
+    [
+      "embodied-agent",
+      66
     ],
     [
       "agent",
       6
     ],
-    [
-      "memory",
-      5
-    ],
     [
       "evaluation",
       4
     ],
-    [
-      "coding-agent",
-      4
-    ],
     [
       "agent-architecture",
       3
     ],
-    [
-      "agent-safety",
-      3
-    ],
     [
       "enterprise-ai",
       3
     ],
-    [
-      "tool-use",
-      2
-    ],
-    [
-      "workflow-agent",
-      1
-    ],
     [
       "llm-infra",
       1
@@ -71,34 +99,6 @@
     [
       "harness",
       1
-    ],
-    [
-      "context-compression",
-      1
-    ],
-    [
-      "agent-productization",
-      1
-    ],
-    [
-      "agent-infra",
-      1
-    ],
-    [
-      "agent-testing",
-      1
-    ],
-    [
-      "harness-engineering",
-      1
-    ],
-    [
-      "human-in-the-loop",
-      1
-    ],
-    [
-      "gui-agent",
-      1
     ]
   ],
   "top_skills": [