diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
new file mode 100644
index 00000000..cc7db8c3
--- /dev/null
+++ b/.github/workflows/ci.yml
@@ -0,0 +1,163 @@
+# SPDX-License-Identifier: Apache-2.0 AND CC-BY-4.0
+# Copyright (c) 2026 NVIDIA Corporation. All rights reserved.
+
+name: Agent Skills Sync Check
+
+on:
+  schedule:
+    - cron: "0 8 * * *" # every day at 08:00 UTC
+  workflow_dispatch:
+  pull_request:
+    branches: [main]
+
+jobs:
+  check-skills:
+    name: Verify skill counts match source repos
+    runs-on: ubuntu-latest
+    permissions:
+      contents: read
+    outputs:
+      drifted: ${{ steps.sync.outputs.drifted }}
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Check skill directories in source repos
+        id: sync
+        env:
+          GH_TOKEN: ${{ secrets.PRIVATE_REPO_ACCESS_TOKEN || secrets.GITHUB_TOKEN }}
+        run: |
+          set -euo pipefail
+
+          DRIFTED=0
+
+          # Format: "display_name|owner/repo|ref|skills_path[+extra_path]|expected_count"
+          REPOS=(
+            "cuOpt|NVIDIA/cuopt|main|skills|19"
+            "TensorRT-LLM|NVIDIA/TensorRT-LLM|main|.claude/skills|4"
+            "Model-Optimizer|NVIDIA/Model-Optimizer|main|.claude/skills|4"
+            "Megatron-Core|NVIDIA/Megatron-LM|main|.claude/skills|2"
+            "Megatron-Bridge|NVIDIA-NeMo/Megatron-Bridge|main|skills|8"
+            "Nemotron Voice Agent|NVIDIA-AI-Blueprints/nemotron-voice-agent|main|.agents/skills|1"
+            "NeMo Gym|NVIDIA-NeMo/Gym|main|.claude/skills|1"
+            "NeMo Evaluator|NVIDIA-NeMo/Evaluator|main|packages/nemo-evaluator-launcher/.claude/skills+packages/nemo-evaluator/.claude/skills|4"
+          )
+
+          REPORT="## Skill Catalog Sync Report\n\n"
+          REPORT+="| Product | Expected | Actual | Status |\n"
+          REPORT+="|---------|----------|--------|--------|\n"
+
+          declare -A UPDATES
+
+          count_skills_in_path() {
+            local repo="$1" ref="$2" path="$3"
+            local result
+            result=$(gh api "repos/${repo}/contents/${path}?ref=${ref}" \
+              --jq '[.[] | select(.type == "dir")] | length' 2>&1)
+            if [ $? -ne 0 ]; then
+              echo "ERR:${result}" 
+            else
+              echo "$result"
+            fi
+          }
+
+          for entry in "${REPOS[@]}"; do
+            IFS='|' read -r label owner_repo ref skills_paths expected <<< "$entry"
+
+            # Sum skill counts across all paths (split on +)
+            actual=0
+            error_msg=""
+            IFS='+' read -ra PATHS <<< "$skills_paths"
+            for spath in "${PATHS[@]}"; do
+              count=$(count_skills_in_path "$owner_repo" "$ref" "$spath")
+              if [[ "$count" == ERR:* ]]; then
+                error_msg="${count#ERR:}"
+                break
+              fi
+              actual=$((actual + count))
+            done
+
+            if [ -n "$error_msg" ]; then
+              REPORT+="| $label | $expected | ERROR | :x: ${error_msg} |\n"
+              echo "::error::${label}: ${error_msg}"
+              DRIFTED=1
+            elif [ "$actual" -eq "$expected" ]; then
+              REPORT+="| $label | $expected | $actual | :white_check_mark: |\n"
+            else
+              REPORT+="| $label | $expected | $actual | :warning: DRIFT |\n"
+              UPDATES["$label"]="$expected:$actual"
+              DRIFTED=1
+            fi
+          done
+
+          echo -e "$REPORT"
+          echo -e "$REPORT" > /tmp/sync-report.md
+          echo "drifted=$DRIFTED" >> "$GITHUB_OUTPUT"
+
+          if [ "$DRIFTED" -ne 0 ]; then
+            echo "::error::Skill counts have drifted — update README.md to match source repos."
+
+            # Persist updates for the fix job
+            for label in "${!UPDATES[@]}"; do
+              echo "${label}=${UPDATES[$label]}"
+            done > /tmp/updates.txt
+          fi
+
+      - name: Upload sync artifacts
+        if: steps.sync.outputs.drifted == '1'
+        uses: actions/upload-artifact@v4
+        with:
+          name: sync-data
+          path: |
+            /tmp/sync-report.md
+            /tmp/updates.txt
+
+      - name: Fail check on drift
+        if: steps.sync.outputs.drifted == '1'
+        run: exit 1
+
+  fix-drift:
+    name: Create PR with corrected skill counts
+    needs: check-skills
+    if: needs.check-skills.outputs.drifted == '1' && github.event_name != 'pull_request'
+    runs-on: ubuntu-latest
+    permissions:
+      contents: write
+      pull-requests: write
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Download sync artifacts
+        uses: actions/download-artifact@v4
+        with:
+          name: sync-data
+          path: /tmp
+
+      - name: Update README.md with corrected counts
+        run: |
+          set -euo pipefail
+
+          while IFS='=' read -r label counts; do
+            IFS=':' read -r old_count new_count <<< "$counts"
+            # Escape spaces in label for sed; match bold product name in table row
+            escaped_label=$(echo "$label" | sed 's/ /[[:space:]]/g')
+            sed -i -E "s/(\*\*${escaped_label}\*\*[^|]*\|[^|]*\| *)${old_count}( *\|)/\1${new_count}\2/" README.md
+          done < /tmp/updates.txt
+
+          if git diff --quiet README.md; then
+            echo "::error::No changes to README.md after update — manual fix may be needed."
+            exit 1
+          fi
+
+          echo "README.md updated:"
+          git diff README.md
+
+      - name: Create pull request
+        uses: peter-evans/create-pull-request@v7
+        with:
+          token: ${{ secrets.GITHUB_TOKEN }}
+          commit-message: "fix: update skill counts to match source repos"
+          branch: auto/sync-skill-counts
+          delete-branch: true
+          title: "fix: sync skill counts with source repos"
+          body-path: /tmp/sync-report.md
+          labels: automated
diff --git a/README.md b/README.md
index c51ae19b..9d728c9f 100644
--- a/README.md
+++ b/README.md
@@ -47,9 +47,9 @@ That's it — the skill activates automatically the next time your agent encount
 |---------|-------------|:------:|-----------------|
 | **cuOpt** | GPU-accelerated optimization — vehicle routing, linear programming, quadratic programming, installation, server deployment, and developer tools. | 19 | [`NVIDIA/cuopt/skills/`](https://github.com/NVIDIA/cuopt/tree/main/skills) |
 | **TensorRT-LLM** | LLM inference optimization — model onboarding to AutoDeploy, CI pipeline failure analysis, and test failure diagnostics. | 4 | [`NVIDIA/TensorRT-LLM/.claude/skills/`](https://github.com/NVIDIA/TensorRT-LLM/tree/main/.claude/skills) |
-| **Model Optimizer** | Model optimization — quantization, sparsity, and distillation for efficient inference. | 4 | [`NVIDIA/Model-Optimizer/.claude/skills/`](https://github.com/NVIDIA/Model-Optimizer/tree/main/.claude/skills) |
-| **Megatron Core** | Large-scale distributed training — model parallelism, pipeline parallelism, and mixed precision. | 2 | [`NVIDIA/Megatron-LM/.claude/skills/`](https://github.com/NVIDIA/Megatron-LM/tree/main/.claude/skills) |
-| **Megatron Bridge** | Bridge between NeMo and Megatron — data processing, model conversion, and training utilities. | 8 | [`NVIDIA-NeMo/Megatron-Bridge/skills/`](https://github.com/NVIDIA-NeMo/Megatron-Bridge/tree/main/skills) |
+| **Model-Optimizer** | Model optimization — quantization, sparsity, and distillation for efficient inference. | 4 | [`NVIDIA/Model-Optimizer/.claude/skills/`](https://github.com/NVIDIA/Model-Optimizer/tree/main/.claude/skills) |
+| **Megatron-Core** | Large-scale distributed training — model parallelism, pipeline parallelism, and mixed precision. | 2 | [`NVIDIA/Megatron-LM/.claude/skills/`](https://github.com/NVIDIA/Megatron-LM/tree/main/.claude/skills) |
+| **Megatron-Bridge** | Bridge between NeMo and Megatron — data processing, model conversion, and training utilities. | 8 | [`NVIDIA-NeMo/Megatron-Bridge/skills/`](https://github.com/NVIDIA-NeMo/Megatron-Bridge/tree/main/skills) |
 | **Nemotron Voice Agent** | Real-time conversational AI — deploy speech-to-speech voice agents on Workstation, Jetson Thor, or Cloud NIMs. | 1 | [`nemotron-voice-agent/.agents/skills/`](https://github.com/NVIDIA-AI-Blueprints/nemotron-voice-agent/tree/main/.agents/skills) |
 | **NeMo Gym** | RL training environments — add benchmarks, resources servers, agent wiring, and reward profiling. | 1 | [`NVIDIA-NeMo/Gym/.claude/skills/`](https://github.com/NVIDIA-NeMo/Gym/tree/main/.claude/skills) |
 | **NeMo Evaluator** | LLM evaluation — launch evaluations, access MLflow results, NeMo Evaluator Launcher assistant, and bring-your-own benchmarks. | 4 | [`NVIDIA-NeMo/Evaluator/.claude/skills/`](https://github.com/NVIDIA-NeMo/Evaluator/tree/main/packages/nemo-evaluator-launcher/.claude/skills) ([+1](https://github.com/NVIDIA-NeMo/Evaluator/tree/main/packages/nemo-evaluator/.claude/skills)) |
@@ -64,9 +64,9 @@ For skill-related issues, feature requests, new skill ideas, discussions, and co
 |---------|--------|-------------|--------------|----------|
 | **cuOpt** | [Issues](https://github.com/NVIDIA/cuopt/issues) | [Discussions](https://github.com/NVIDIA/cuopt/discussions) | [Contributing](https://github.com/NVIDIA/cuopt/blob/main/CONTRIBUTING.md) | [Security](https://github.com/NVIDIA/cuopt/blob/main/SECURITY.md) |
 | **TensorRT-LLM** | [Issues](https://github.com/NVIDIA/TensorRT-LLM/issues) | [Discussions](https://github.com/NVIDIA/TensorRT-LLM/discussions) | [Contributing](https://github.com/NVIDIA/TensorRT-LLM/blob/main/CONTRIBUTING.md) | [Security](https://github.com/NVIDIA/TensorRT-LLM/blob/main/SECURITY.md) |
-| **Model Optimizer** | [Issues](https://github.com/NVIDIA/Model-Optimizer/issues) | — | [Contributing](https://github.com/NVIDIA/Model-Optimizer/blob/main/CONTRIBUTING.md) | [Security](https://github.com/NVIDIA/Model-Optimizer/blob/main/SECURITY.md) |
-| **Megatron Core** | [Issues](https://github.com/NVIDIA/Megatron-LM/issues) | [Discussions](https://github.com/NVIDIA/Megatron-LM/discussions) | [Contributing](https://github.com/NVIDIA/Megatron-LM/blob/main/CONTRIBUTING.md) | — |
-| **Megatron Bridge** | [Issues](https://github.com/NVIDIA-NeMo/Megatron-Bridge/issues) | [Discussions](https://github.com/NVIDIA-NeMo/Megatron-Bridge/discussions) | [Contributing](https://github.com/NVIDIA-NeMo/Megatron-Bridge/blob/main/CONTRIBUTING.md) | — |
+| **Model-Optimizer** | [Issues](https://github.com/NVIDIA/Model-Optimizer/issues) | — | [Contributing](https://github.com/NVIDIA/Model-Optimizer/blob/main/CONTRIBUTING.md) | [Security](https://github.com/NVIDIA/Model-Optimizer/blob/main/SECURITY.md) |
+| **Megatron-Core** | [Issues](https://github.com/NVIDIA/Megatron-LM/issues) | [Discussions](https://github.com/NVIDIA/Megatron-LM/discussions) | [Contributing](https://github.com/NVIDIA/Megatron-LM/blob/main/CONTRIBUTING.md) | — |
+| **Megatron-Bridge** | [Issues](https://github.com/NVIDIA-NeMo/Megatron-Bridge/issues) | [Discussions](https://github.com/NVIDIA-NeMo/Megatron-Bridge/discussions) | [Contributing](https://github.com/NVIDIA-NeMo/Megatron-Bridge/blob/main/CONTRIBUTING.md) | — |
 | **Nemotron Voice Agent** | [Issues](https://github.com/NVIDIA-AI-Blueprints/nemotron-voice-agent/issues) | [Discussions](https://github.com/NVIDIA-AI-Blueprints/nemotron-voice-agent/discussions) | [Contributing](https://github.com/NVIDIA-AI-Blueprints/nemotron-voice-agent/blob/main/CONTRIBUTING.md) | [Security](https://github.com/NVIDIA-AI-Blueprints/nemotron-voice-agent/blob/main/SECURITY.md) |
 | **NeMo Gym** | [Issues](https://github.com/NVIDIA-NeMo/Gym/issues) | [Discussions](https://github.com/NVIDIA-NeMo/Gym/discussions) | [Contributing](https://github.com/NVIDIA-NeMo/Gym/blob/main/CONTRIBUTING.md) | [Security](https://github.com/NVIDIA-NeMo/Gym/blob/main/SECURITY.md) |
 | **NeMo Evaluator** | [Issues](https://github.com/NVIDIA-NeMo/Evaluator/issues) | [Discussions](https://github.com/NVIDIA-NeMo/Evaluator/discussions) | [Contributing](https://github.com/NVIDIA-NeMo/Evaluator/blob/main/CONTRIBUTING.md) | [Security](https://github.com/NVIDIA-NeMo/Evaluator/blob/main/SECURITY.md) |