diff --git a/.github/scripts/format_benchstat.py b/.github/scripts/format_benchstat.py
new file mode 100755
index 0000000..44288fe
--- /dev/null
+++ b/.github/scripts/format_benchstat.py
@@ -0,0 +1,170 @@
+#!/usr/bin/env python3
+"""
+Formats raw Go benchstat text output into a clean, executive GitHub Step Summary Markdown report.
+"""
+import sys
+import re
+
+def parse_val_unit(s):
+    """Parses value string like '59.30m', '7.601Mi', '18.84k' into a numeric float."""
+    s = s.strip()
+    m = re.match(r'^([0-9.]+)([a-zA-Z%]*)$', s)
+    if not m:
+        return 0.0
+    val_str, unit = m.groups()
+    val = float(val_str)
+    multiplier = 1.0
+    if unit == 'k': multiplier = 1e3
+    elif unit in ('M', 'Mi'): multiplier = 1e6
+    elif unit in ('G', 'Gi'): multiplier = 1e9
+    elif unit == 'm': multiplier = 1e-3
+    elif unit in ('u', 'µ'): multiplier = 1e-6
+    elif unit == 'n': multiplier = 1e-9
+    return val * multiplier
+
+def compute_pct(old_s, new_s):
+    old_v = parse_val_unit(old_s)
+    new_v = parse_val_unit(new_s)
+    if old_v == 0:
+        return 0.0
+    return ((new_v - old_v) / old_v) * 100.0
+
+def format_benchstat(text):
+    lines = text.splitlines()
+    current_metric = None
+    geomeans = {}
+    fixtures = {} # fix_name -> {'time': (old, new), 'mem': (old, new), 'allocs': (old, new)}
+
+    for line in lines:
+        line_str = line.strip()
+        if 'sec/op' in line_str and 'vs base' in line_str:
+            current_metric = 'time'
+        elif 'B/op' in line_str and 'vs base' in line_str:
+            current_metric = 'mem'
+        elif 'allocs/op' in line_str and 'vs base' in line_str:
+            current_metric = 'allocs'
+        elif line_str.startswith('geomean'):
+            parts = line_str.split()
+            if len(parts) >= 3 and current_metric:
+                delta = parts[3] if len(parts) >= 4 else "~"
+                geomeans[current_metric] = (parts[1], parts[2], delta)
+        elif line_str.startswith('Pipeline/'):
+            parts = line_str.split()
+            if len(parts) >= 3 and current_metric:
+                fixture_raw = parts[0]
+                fix_name = fixture_raw.replace('Pipeline/', '')
+                if '-' in fix_name:
+                    fix_name = fix_name.rsplit('-', 1)[0]
+                
+                vals = [p for p in parts[1:] if re.match(r'^[0-9.]+[a-zA-Z]*$', p)]
+                if len(vals) >= 2:
+                    old_v, new_v = vals[0], vals[1]
+                    if fix_name not in fixtures:
+                        fixtures[fix_name] = {}
+                    fixtures[fix_name][current_metric] = (old_v, new_v)
+
+    # Build Markdown Output
+    md = []
+    md.append("### 📊 Benchmark A/B Comparison Report")
+    md.append("")
+    md.append("#### 📈 Key Metrics Summary (Geometric Mean)")
+    md.append("| Metric | Baseline | Current | Delta | Status |")
+    md.append("| :--- | :---: | :---: | :---: | :---: |")
+
+    metric_meta = {
+        'time': ('Execution Time', 'sec/op'),
+        'mem': ('Memory Allocated', 'B/op'),
+        'allocs': ('Allocations', 'allocs/op')
+    }
+
+    for key in ['time', 'mem', 'allocs']:
+        if key in geomeans:
+            old_g, new_g, _ = geomeans[key]
+            pct = compute_pct(old_g, new_g)
+            name, _ = metric_meta[key]
+            
+            if abs(pct) < 0.5:
+                status = "➖ Neutral"
+                badge = f"`{pct:+.2f}%`"
+            elif pct < 0:
+                status = "🟢 Faster" if key == 'time' else "🟢 Reduced"
+                badge = f"**`{pct:+.2f}%`**"
+            else:
+                status = "🔴 Slower" if key == 'time' else "🟡 Increased"
+                badge = f"**`{pct:+.2f}%`**"
+                
+            md.append(f"| **{name}** | `{old_g}` | `{new_g}` | {badge} | {status} |")
+
+    md.append("")
+
+    # Sort changes by time delta
+    changes = []
+    for fix_name, data in fixtures.items():
+        if 'time' in data:
+            old_t, new_t = data['time']
+            pct = compute_pct(old_t, new_t)
+            changes.append((fix_name, pct, old_t, new_t, data.get('mem', ('-', '-'))))
+
+    changes.sort(key=lambda x: x[1])
+
+    speedups = [c for c in changes if c[1] <= -10.0]
+    regressions = [c for c in changes if c[1] >= 10.0]
+
+    if speedups or regressions:
+        md.append("#### ⚡ Notable Performance Shifts (≥ 10% change)")
+        md.append("| Fixture | Baseline Time | Current Time | Time Delta | Memory Delta |")
+        md.append("| :--- | :---: | :---: | :---: | :---: |")
+        
+        for fix_name, pct, old_t, new_t, mem_tuple in speedups + regressions:
+            icon = "🚀" if pct < 0 else "⚠️"
+            old_m, new_m = mem_tuple
+            mem_pct_str = "-"
+            if old_m != '-' and new_m != '-':
+                m_pct = compute_pct(old_m, new_m)
+                mem_pct_str = f"`{m_pct:+.1f}%`"
+            md.append(f"| {icon} `{fix_name}` | `{old_t}` | `{new_t}` | **`{pct:+.1f}%`** | {mem_pct_str} |")
+        md.append("")
+
+    # Collapsible Table for All Fixtures
+    md.append("<details>")
+    md.append(f"<summary><b>🔍 Full Breakdown ({len(fixtures)} Fixtures)</b></summary>")
+    md.append("")
+    md.append("| Fixture | Baseline Time | Current Time | Time Delta | Baseline Mem | Current Mem | Mem Delta |")
+    md.append("| :--- | :---: | :---: | :---: | :---: | :---: | :---: |")
+
+    for fix_name, data in sorted(fixtures.items()):
+        old_t, new_t = data.get('time', ('-', '-'))
+        old_m, new_m = data.get('mem', ('-', '-'))
+        
+        t_pct_str = "`~`"
+        if old_t != '-' and new_t != '-':
+            t_pct = compute_pct(old_t, new_t)
+            t_pct_str = f"`{t_pct:+.1f}%`" if abs(t_pct) >= 0.5 else "`~`"
+            
+        m_pct_str = "`~`"
+        if old_m != '-' and new_m != '-':
+            m_pct = compute_pct(old_m, new_m)
+            m_pct_str = f"`{m_pct:+.1f}%`" if abs(m_pct) >= 0.5 else "`~`"
+
+        md.append(f"| `{fix_name}` | `{old_t}` | `{new_t}` | {t_pct_str} | `{old_m}` | `{new_m}` | {m_pct_str} |")
+
+    md.append("")
+    md.append("</details>")
+    md.append("")
+    md.append("<details>")
+    md.append("<summary><b>📄 Raw benchstat Console Output</b></summary>")
+    md.append("")
+    md.append("```")
+    md.append(text.strip())
+    md.append("```")
+    md.append("</details>")
+
+    return "\n".join(md)
+
+if __name__ == '__main__':
+    if len(sys.argv) > 1:
+        with open(sys.argv[1], 'r') as f:
+            content = f.read()
+    else:
+        content = sys.stdin.read()
+    print(format_benchstat(content))
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
index a2598e0..ac1c2ec 100644
--- a/.github/workflows/ci.yml
+++ b/.github/workflows/ci.yml
@@ -17,13 +17,9 @@ concurrency:
   cancel-in-progress: ${{ github.ref != 'refs/heads/main' }}
 
 jobs:
-  test:
-    name: Test & Build
-    runs-on: ${{ matrix.os }}
-    strategy:
-      fail-fast: false
-      matrix:
-        os: [ubuntu-latest, macos-latest, windows-latest]
+  lint:
+    name: Lint & Format Check
+    runs-on: ubuntu-latest
 
     steps:
       - name: Checkout code
@@ -37,40 +33,88 @@ jobs:
           cache-dependency-path: 'go.sum'
 
       - name: Lint (golangci-lint)
-        if: matrix.os == 'ubuntu-latest'
         uses: golangci/golangci-lint-action@v9
         with:
           version: v2.12
           args: --timeout=5m
 
       - name: Verify formatting (golangci-lint fmt)
-        if: matrix.os == 'ubuntu-latest'
         run: golangci-lint fmt --diff ./...
 
+  test-linux:
+    name: Test & Coverage (Linux)
+    needs: lint
+    runs-on: ubuntu-latest
+
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+
+      - name: Set up Go
+        uses: actions/setup-go@v5
+        with:
+          go-version-file: 'go.mod'
+          cache: true
+          cache-dependency-path: 'go.sum'
+
       - name: Install gotestsum
         run: go install gotest.tools/gotestsum@latest
 
-      - name: Run test suite
-        if: matrix.os != 'ubuntu-latest'
-        run: gotestsum --format pkgname --junitfile unit-tests.xml -- ./...
-
       - name: Run test suite with coverage
-        if: matrix.os == 'ubuntu-latest'
         run: >-
           gotestsum --format pkgname --junitfile unit-tests.xml --
           -coverprofile=coverage.out -covermode=atomic -coverpkg=./...
           ./...
 
-      - name: Print coverage summary
-        if: matrix.os == 'ubuntu-latest'
-        run: go tool cover -func=coverage.out | tail -1
+      - name: Print & publish coverage summary
+        run: |
+          go tool cover -func=coverage.out | tail -1
+          echo "### 🧪 Code Coverage" >> $GITHUB_STEP_SUMMARY
+          echo '```' >> $GITHUB_STEP_SUMMARY
+          go tool cover -func=coverage.out | tail -15 >> $GITHUB_STEP_SUMMARY
+          echo '```' >> $GITHUB_STEP_SUMMARY
+          go tool cover -html=coverage.out -o coverage.html
 
       - name: Upload coverage report
-        if: matrix.os == 'ubuntu-latest'
         uses: actions/upload-artifact@v4
         with:
           name: coverage-report
-          path: coverage.out
+          path: |
+            coverage.out
+            coverage.html
+
+      - name: Publish Test Report
+        uses: mikepenz/action-junit-report@v4
+        if: always()
+        with:
+          report_paths: '**/unit-tests.xml'
+          check_name: 'Test Report (ubuntu-latest)'
+
+  test-cross:
+    name: Cross-Platform Test
+    needs: lint
+    runs-on: ${{ matrix.os }}
+    strategy:
+      fail-fast: false
+      matrix:
+        os: [macos-latest, windows-latest]
+
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+
+      - name: Set up Go
+        uses: actions/setup-go@v5
+        with:
+          go-version-file: 'go.mod'
+          cache: true
+          cache-dependency-path: 'go.sum'
+
+      - name: Install gotestsum
+        run: go install gotest.tools/gotestsum@latest
+
+      - name: Run test suite
+        run: gotestsum --format pkgname --junitfile unit-tests.xml -- ./...
 
       - name: Publish Test Report
         uses: mikepenz/action-junit-report@v4
@@ -79,26 +123,64 @@ jobs:
           report_paths: '**/unit-tests.xml'
           check_name: 'Test Report (${{ matrix.os }})'
 
-      - name: Benchmark smoke test
-        if: matrix.os == 'ubuntu-latest'
-        run: go test ./tests/integration/ -bench=BenchmarkPipeline -benchmem -run=^$ -count=1 -benchtime=1x
+  benchmark:
+    name: Benchmark A/B Comparison
+    needs: [lint, test-linux]
+    runs-on: ubuntu-latest
+
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+
+      - name: Check performance-critical file changes
+        id: filter
+        run: |
+          BASE_REF="${{ github.event.pull_request.base.sha || 'HEAD~1' }}"
+          CHANGED=$(git diff --name-only $BASE_REF HEAD || true)
+          echo "Changed files:"
+          echo "$CHANGED"
+          if echo "$CHANGED" | grep -qE '^(internal/engine/|internal/treesitter/|internal/actions/|tests/integration/|go\.mod|go\.sum)'; then
+            echo "::notice::Performance-critical files changed. Running A/B benchmarks..."
+            echo "run_benchmarks=true" >> $GITHUB_OUTPUT
+          else
+            echo "::notice::No performance-critical files changed. Skipping benchmarks."
+            echo "run_benchmarks=false" >> $GITHUB_OUTPUT
+          fi
+
+      - name: Set up Go
+        if: steps.filter.outputs.run_benchmarks == 'true'
+        uses: actions/setup-go@v5
+        with:
+          go-version-file: 'go.mod'
+          cache: true
+          cache-dependency-path: 'go.sum'
+
+      - name: Install benchstat
+        if: steps.filter.outputs.run_benchmarks == 'true'
+        run: go install golang.org/x/perf/cmd/benchstat@latest
+
+      - name: Run benchmarks on current HEAD
+        if: steps.filter.outputs.run_benchmarks == 'true'
+        run: go test ./tests/integration/ -bench=BenchmarkPipeline -benchmem -run=^$ -count=3 > bench_new.txt
 
-      - name: Benchmark A/B comparison (benchstat & GITHUB_STEP_SUMMARY)
-        if: matrix.os == 'ubuntu-latest'
+      - name: Fetch baseline branch
+        if: steps.filter.outputs.run_benchmarks == 'true'
         run: |
-          go install golang.org/x/perf/cmd/benchstat@latest
-          echo "==> Running benchmarks on current HEAD..."
-          go test ./tests/integration/ -bench=BenchmarkPipeline -benchmem -run=^$ -count=3 > bench_new.txt
-          echo "==> Fetching baseline..."
-          git fetch origin ${{ github.event.pull_request.base.ref || 'main' }} --depth=2 || true
           BASE_REF="${{ github.event.pull_request.base.sha || 'HEAD~1' }}"
           git checkout $BASE_REF || true
-          echo "==> Running benchmarks on baseline ($BASE_REF)..."
+
+      - name: Run benchmarks on baseline
+        if: steps.filter.outputs.run_benchmarks == 'true'
+        run: |
           go test ./tests/integration/ -bench=BenchmarkPipeline -benchmem -run=^$ -count=3 > bench_old.txt || true
-          
-          # Render Markdown comparison table for GitHub Actions Summary Tab
-          echo "### 📊 Benchmark A/B Comparison vs Baseline" >> $GITHUB_STEP_SUMMARY
-          echo '```' >> $GITHUB_STEP_SUMMARY
-          benchstat bench_old.txt bench_new.txt >> $GITHUB_STEP_SUMMARY || true
-          echo '```' >> $GITHUB_STEP_SUMMARY
-          rm -f bench_old.txt bench_new.txt
+          echo "==> Restoring PR HEAD commit to access scripts..."
+          git checkout ${{ github.sha }} || git checkout - || true
+
+      - name: Render Benchmark A/B Summary
+        if: steps.filter.outputs.run_benchmarks == 'true'
+        run: |
+          benchstat bench_old.txt bench_new.txt > benchstat_raw.txt || true
+          python3 .github/scripts/format_benchstat.py benchstat_raw.txt >> $GITHUB_STEP_SUMMARY || cat benchstat_raw.txt >> $GITHUB_STEP_SUMMARY
+          rm -f bench_old.txt bench_new.txt benchstat_raw.txt
diff --git a/go.mod b/go.mod
index 9a297f3..e023afb 100644
--- a/go.mod
+++ b/go.mod
@@ -7,7 +7,7 @@ require (
 	github.com/charmbracelet/bubbletea v1.3.10
 	github.com/charmbracelet/lipgloss v1.1.0
 	github.com/mattn/go-isatty v0.0.20
-	github.com/odvcencio/gotreesitter v0.16.0
+	github.com/odvcencio/gotreesitter v0.48.0
 	github.com/spf13/cobra v1.10.2
 	gopkg.in/yaml.v3 v3.0.1
 )
diff --git a/go.sum b/go.sum
index 2ceef3c..e47f433 100644
--- a/go.sum
+++ b/go.sum
@@ -46,8 +46,8 @@ github.com/muesli/cancelreader v0.2.2 h1:3I4Kt4BQjOR54NavqnDogx/MIoWBFa0StPA8ELU
 github.com/muesli/cancelreader v0.2.2/go.mod h1:3XuTXfFS2VjM+HTLZY9Ak0l6eUKfijIfMUZ4EgX0QYo=
 github.com/muesli/termenv v0.16.0 h1:S5AlUN9dENB57rsbnkPyfdGuWIlkmzJjbFf0Tf5FWUc=
 github.com/muesli/termenv v0.16.0/go.mod h1:ZRfOIKPFDYQoDFF4Olj7/QJbW60Ol/kL1pU3VfY/Cnk=
-github.com/odvcencio/gotreesitter v0.16.0 h1:pg29HG6idOpmbh6BkPc6ncFP0C8rH9ISAzFVNdHm2ss=
-github.com/odvcencio/gotreesitter v0.16.0/go.mod h1:ccYZsDUmAJQAtliLsNHT33F3X4AN7f/Z6JGiPNZoEzY=
+github.com/odvcencio/gotreesitter v0.48.0 h1:Gr9XVbGs/1yqAsLZWp2zmW91XM4YEMJsWiZa35fm4R8=
+github.com/odvcencio/gotreesitter v0.48.0/go.mod h1:hBVkghd0paaYAVwd2087vfwdeU984bQbMo9LvpE0moo=
 github.com/rivo/uniseg v0.4.7 h1:WUdvkW8uEhrYfLC4ZzdpI2ztxP1I582+49Oc5Mq64VQ=
 github.com/rivo/uniseg v0.4.7/go.mod h1:FN3SvrM+Zdj16jyLfmOkMNblXMcoc8DfTHruCPUcx88=
 github.com/russross/blackfriday/v2 v2.1.0/go.mod h1:+Rmxgy9KzJVeS9/2gXHxylqXiyQDYRxCVz55jmeOWTM=