diff --git a/.gitea/workflows/adversarial-review.yml b/.gitea/workflows/adversarial-review.yml
index 9dd178e..4c8e428 100644
--- a/.gitea/workflows/adversarial-review.yml
+++ b/.gitea/workflows/adversarial-review.yml
@@ -44,7 +44,7 @@ jobs:
     # 3 cloud models, all concurrent, 3-lens suite. ~12 min typical.
     timeout-minutes: 30
     steps:
-      - uses: docker://gitea.stevedudenhoeffer.com/steve/gadfly:sha-6e3a83c
+      - uses: docker://gitea.stevedudenhoeffer.com/steve/gadfly:sha-d0de034
         env:
           GITEA_API: ${{ github.server_url }}/api/v1/repos/${{ github.repository }}
           GITEA_TOKEN: ${{ secrets.GITEA_TOKEN }}
@@ -53,9 +53,14 @@ jobs:
           # on a P2-review measurement they took 26–29 min (with lens timeouts)
           # and contributed ZERO real findings — the two cloud models found every
           # genuine bug in 6–12 min. Cloud-only is faster AND higher-signal.
-          # 3 cloud models, one consolidated comment each, all run in parallel.
+          # 3 cloud models. Concurrency now lives in the LENSES, not the models:
+          # one model runs at a time (PROVIDER_CONCURRENCY=1) with its 3 lenses
+          # concurrent (PROVIDER_LENS_CONCURRENCY=3). So the first model's
+          # comment lands sooner and each model finishes a bit faster, at the
+          # cost of the other two models' comments arriving in series after it.
           GADFLY_MODELS: "minimax-m3:cloud,deepseek-v4-flash:cloud,glm-5.2:cloud"
-          GADFLY_PROVIDER_CONCURRENCY: "ollama-cloud=3"
+          GADFLY_PROVIDER_CONCURRENCY: "ollama-cloud=1"
+          GADFLY_PROVIDER_LENS_CONCURRENCY: "ollama-cloud=3"
           # Default => the 3-lens suite (security, correctness, error-handling).
           # Set the repo var GADFLY_SPECIALISTS to override (csv / "all" / "auto").
           GADFLY_SPECIALISTS: ${{ vars.GADFLY_SPECIALISTS || 'security,correctness,error-handling' }}