Fix AI filtering bugs, add special award shortlist integration

Part 1 - Bug Fixes:
- Fix toProjectWithRelations() stripping file fields needed by AI (detectedLang, textContent, etc.)
- Fix parseAIData() reading flat when aiScreeningJson is nested under rule ID
- Fix getAIConfidenceScore() with same nesting issue (always returned 0)

Part 2 - Special Award Track Integration:
- Add shortlistSize to SpecialAward, qualityScore/shortlisted/confirmed fields to AwardEligibility
- Add specialAwardId to Round for award-owned rounds
- Update AI eligibility service to return qualityScore (0-100) for ranking
- Update eligibility job with filteringRoundId scoping and auto-shortlist top N
- Add 8 new specialAward router procedures (listForRound, runEligibilityForRound,
  listShortlist, toggleShortlisted, confirmShortlist, listRounds, createRound, deleteRound)
- Create award-shortlist.tsx component with ranked table, shortlist checkboxes, confirm dialog
- Add "Special Award Tracks" section to filtering dashboard

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

src/server/services/ai-award-eligibility.ts

@@ -48,6 +48,7 @@ Return a JSON object:
       "project_id": "PROJECT_001",
       "eligible": true/false,
       "confidence": 0.0-1.0,
+      "quality_score": 0-100,
       "reasoning": "2-3 sentence explanation covering key dimensions",
       "dimensionScores": {
         "geographic": 0.0-1.0,
@@ -59,6 +60,8 @@ Return a JSON object:
   ]
 }
 
+quality_score is a 0-100 integer measuring how well the project fits the award criteria (used for ranking shortlists). 100 = perfect fit, 0 = no fit. Even ineligible projects should receive a score for reference.
+
 ## Guidelines
 - Base evaluation only on provided data — do not infer missing information
 - eligible=true only when ALL required dimensions score above 0.5
@@ -77,6 +80,7 @@ export interface EligibilityResult {
   projectId: string
   eligible: boolean
   confidence: number
+  qualityScore: number
   reasoning: string
   method: 'AUTO' | 'AI'
 }
@@ -229,6 +233,7 @@ Evaluate eligibility for each project.`
         project_id: string
         eligible: boolean
         confidence: number
+        quality_score?: number
         reasoning: string
       }>
     }
@@ -273,6 +278,7 @@ Evaluate eligibility for each project.`
           projectId: mapping.realId,
           eligible: eval_.eligible,
           confidence: eval_.confidence,
+          qualityScore: Math.max(0, Math.min(100, eval_.quality_score ?? 0)),
           reasoning: eval_.reasoning,
           method: 'AI',
         })
@@ -305,6 +311,7 @@ Evaluate eligibility for each project.`
           projectId: mapping.realId,
           eligible: false,
           confidence: 0,
+          qualityScore: 0,
           reasoning: 'AI response parse error — requires manual review',
           method: 'AI',
         })
@@ -333,6 +340,7 @@ export async function aiInterpretCriteria(
         projectId: p.id,
         eligible: false,
         confidence: 0,
+        qualityScore: 0,
         reasoning: 'AI unavailable — requires manual eligibility review',
         method: 'AI' as const,
       }))
@@ -401,6 +409,7 @@ export async function aiInterpretCriteria(
       projectId: p.id,
       eligible: false,
       confidence: 0,
+      qualityScore: 0,
       reasoning: `AI error: ${classified.message}`,
       method: 'AI' as const,
     }))