feat: robust metric extraction with confidence score and proof snippets

- fixed Year-Prefix Bug in MetricParser
- added metric_confidence and metric_proof_text to database
- added Entity-Check and Annual-Priority to LLM prompt
- improved UI: added confidence traffic light and mouse-over proof tooltip
- restored missing API endpoints (create, bulk, wiki-override)

company-explorer/frontend/node_modules/tailwindcss/src/lib/sharedState.js

@@ -0,0 +1,57 @@
+export const env =
+  typeof process !== 'undefined'
+    ? {
+        NODE_ENV: process.env.NODE_ENV,
+        DEBUG: resolveDebug(process.env.DEBUG),
+      }
+    : {
+        NODE_ENV: 'production',
+        DEBUG: false,
+      }
+
+export const contextMap = new Map()
+export const configContextMap = new Map()
+export const contextSourcesMap = new Map()
+export const sourceHashMap = new Map()
+export const NOT_ON_DEMAND = new String('*')
+
+export const NONE = Symbol('__NONE__')
+
+export function resolveDebug(debug) {
+  if (debug === undefined) {
+    return false
+  }
+
+  // Environment variables are strings, so convert to boolean
+  if (debug === 'true' || debug === '1') {
+    return true
+  }
+
+  if (debug === 'false' || debug === '0') {
+    return false
+  }
+
+  // Keep the debug convention into account:
+  // DEBUG=* -> This enables all debug modes
+  // DEBUG=projectA,projectB,projectC -> This enables debug for projectA, projectB and projectC
+  // DEBUG=projectA:* -> This enables all debug modes for projectA (if you have sub-types)
+  // DEBUG=projectA,-projectB -> This enables debug for projectA and explicitly disables it for projectB
+
+  if (debug === '*') {
+    return true
+  }
+
+  let debuggers = debug.split(',').map((d) => d.split(':')[0])
+
+  // Ignoring tailwindcss
+  if (debuggers.includes('-tailwindcss')) {
+    return false
+  }
+
+  // Including tailwindcss
+  if (debuggers.includes('tailwindcss')) {
+    return true
+  }
+
+  return false
+}