Replace "effective tokens" with "AI Credits" in user-facing text

.github/workflows/daily-agentrx-trace-optimizer.md

@@ -16,7 +16,7 @@ experiments:
     description: "Test whether delegating trajectory-builder, artifacts-summarizer, and failure-pattern-classifier to small-model sub-agents improves recommendation quality vs. inline analysis by the main agent"
     hypothesis: "H0: no change in issue quality or run success rate. H1: sub_agents variant yields higher evidence completeness score with equal or lower token cost"
     metric: issue_evidence_completeness
-    secondary_metrics: [run_success_rate, effective_tokens_total, run_duration_ms]
+    secondary_metrics: [run_success_rate, ai_credits_total, run_duration_ms]
     guardrail_metrics:
       - name: empty_output_rate
         threshold: "<=0.10"

.github/workflows/daily-cache-strategy-analyzer.md

@@ -21,8 +21,8 @@ experiments:
   model_size:
     variants: [gpt-5.4, gpt-5-codex]
     description: "Compares codex-compatible models for cache issue detection quality and efficiency."
-    hypothesis: "H0: no change in issue creation rate or run success rate. H1: gpt-5-codex reduces effective tokens while keeping run success rate >=0.90."
-    metric: effective_tokens_total
+    hypothesis: "H0: no change in issue creation rate or run success rate. H1: gpt-5-codex reduces AI Credits while keeping run success rate >=0.90."
+    metric: ai_credits_total
     secondary_metrics: [run_success_rate, run_duration_ms]
     guardrail_metrics:
       - name: run_success_rate

.github/workflows/daily-caveman-optimizer.md

@@ -23,7 +23,7 @@ experiments:
     variants: [claude-sonnet-4.6, claude-haiku-4.5]
     description: "Tests whether Claude Haiku produces equivalent instruction conciseness improvements at lower token cost versus Claude Sonnet."
     hypothesis: "H0: no change in PR creation rate or run success rate. H1: Claude Haiku reduces AI credit usage >=30% with equivalent run success rate (>=0.90)."
-    metric: effective_tokens_total
+    metric: ai_credits_total
     secondary_metrics: [run_success_rate, run_duration_ms]
     guardrail_metrics:
       - name: run_success_rate