diff --git a/scripts/rugguard.py b/scripts/rugguard.py
index dcd51cc..1536bd0 100644
--- a/scripts/rugguard.py
+++ b/scripts/rugguard.py
@@ -1627,7 +1627,132 @@ def format_markdown(report: RugReport) -> str:
 def format_json(report: RugReport) -> str:
     """Format report as pretty JSON."""
     return json.dumps(report.to_dict(), indent=2, default=str)
-# ── CLI Entry Point ────────────────────────────────────────────────────────
+
+
+# ── Comparison Table ──────────────────────────────────────────────────────
+
+
+def _format_comparison_table(reports: list[RugReport],
+                             sort_by: str = "score") -> str:
+    """Render a side-by-side ASCII comparison table for multiple tokens.
+
+    Each column = one token, auto-sized to widest value.
+    Tokens are sorted by safety_score ascending (riskiest first).
+    """
+    if not reports:
+        return ""
+
+    # Sort: score asc (riskiest first)
+    if sort_by == "score":
+        key_fn = lambda r: r.safety_score
+    elif sort_by == "name":
+        key_fn = lambda r: (r.token.name or r.token.symbol or r.token.address)
+    elif sort_by == "age":
+        key_fn = lambda r: -r.score.age_risk  # less age risk = older
+    elif sort_by == "liquidity":
+        key_fn = lambda r: -(r.dex_data.get("liquidity_usd", 0) if r.dex_data else 0)
+    else:
+        key_fn = lambda r: r.safety_score
+    sorted_reports = sorted(reports, key=key_fn)
+
+    # Build rows: each is a list of values, one per token
+    headers = ["Metric"]
+    for r in sorted_reports:
+        sym = r.token.symbol or r.token.name or r.token.address[:8]
+        headers.append(sym)
+
+    rows: list[list[str]] = [headers]
+
+    def val(name: str, vals: list[str]) -> None:
+        rows.append([name] + vals)
+
+    val("Safety Score", [str(r.safety_score) for r in sorted_reports])
+    val("Risk Level", [r.risk_level for r in sorted_reports])
+
+    # Price
+    prices = []
+    for r in sorted_reports:
+        if r.dex_data and r.dex_data.get("price_usd"):
+            p = r.dex_data["price_usd"]
+            prices.append(f"${p:.8f}" if p < 1 else f"${p:.4f}")
+        else:
+            prices.append("—")
+    val("Price", prices)
+
+    # 24h Change
+    changes = []
+    for r in sorted_reports:
+        if r.dex_data and r.dex_data.get("price_change_24h"):
+            pct = r.dex_data["price_change_24h"]
+            changes.append(f"{pct:+.2f}%")
+        else:
+            changes.append("—")
+    val("24h Change", changes)
+
+    # Liquidity
+    liqs = []
+    for r in sorted_reports:
+        if r.dex_data and r.dex_data.get("liquidity_usd"):
+            liqs.append(f"${r.dex_data['liquidity_usd']:,.0f}")
+        else:
+            liqs.append("—")
+    val("Liquidity", liqs)
+
+    # Volume 24h
+    vols = []
+    for r in sorted_reports:
+        if r.dex_data and r.dex_data.get("volume_24h"):
+            vols.append(f"${r.dex_data['volume_24h']:,.0f}")
+        else:
+            vols.append("—")
+    val("Volume 24h", vols)
+
+    # Holders
+    hldrs = []
+    for r in sorted_reports:
+        if r.holders:
+            hldrs.append(str(r.holders.total_holders))
+        else:
+            hldrs.append("—")
+    val("Holders", hldrs)
+
+    # Top 10 %
+    top10 = []
+    for r in sorted_reports:
+        if r.holders:
+            top10.append(f"{r.holders.top_10_pct:.1f}%")
+        else:
+            top10.append("—")
+    val("Top 10%", top10)
+
+    # Warnings count
+    wcounts = [str(len(r.warnings)) for r in sorted_reports]
+    val("Warnings", wcounts)
+
+    # Auto-size column widths
+    col_widths: list[int] = []
+    for ci in range(len(rows[0])):
+        col_widths.append(max(len(r[ci]) for r in rows))
+    col_widths[0] = max(col_widths[0], 12)  # min metric column width
+
+    # Render table
+    sep = " | "
+    lines = []
+    # Header
+    hdr_parts = [h.ljust(col_widths[i]) for i, h in enumerate(rows[0])]
+    lines.append(sep.join(hdr_parts))
+    # Separator
+    sep_parts = ["-" * col_widths[i] for i in range(len(col_widths))]
+    lines.append(sep.join(sep_parts))
+    # Rows
+    for r in rows[1:]:
+        parts = [r[i].ljust(col_widths[i]) for i in range(len(r))]
+        lines.append(sep.join(parts))
+
+    return "\n".join(lines)
+
+
+# ── Export Helpers (CSV / JSONL) ──────────────────────────────────────────
 
 def cli_token(args: list[str]) -> None:
     mint = args[0] if args else ""
@@ -1660,12 +1785,67 @@ def cli_wallet(args: list[str]) -> None:
 
     if result.get("risky_count", 0) > 0:
         sys.exit(2)
+
+
+def cli_compare(args: list[str]) -> None:
+    """Compare multiple tokens side-by-side."""
+    if not args:
+        print('Usage: python rugguard.py compare <MINT1> <MINT2> [<MINT3> ...]', file=sys.stderr)
+        sys.exit(1)
+
+    # Parse flags
+    as_json = "--json" in args
+    sort_by = "score"
+    mints = []
+    for a in args:
+        if a == "--json":
+            continue
+        if a.startswith("--sort="):
+            sort_by = a.split("=", 1)[1]
+        elif a == "--sort" and args.index(a) + 1 < len(args):
+            continue  # handled by next arg
+        else:
+            mints.append(a)
+
+    if len(mints) < 2:
+        print("Error: need at least 2 mint addresses to compare", file=sys.stderr)
+        sys.exit(1)
+
+    reports: list[RugReport] = []
+    errors: list[str] = []
+
+    for mint in mints:
+        try:
+            report = rug_check_token(mint.strip())
+            reports.append(report)
+        except Exception as e:
+            errors.append(f"{mint[:8]}...: {e}")
+            # Insert a minimal placeholder for failed token
+            reports.append(RugReport(
+                token=TokenMeta(address=mint),
+                safety_score=0, risk_level="ERROR",
+                score=RugScore(), flags=RugFlags(),
+                warnings=[], recommendation="Check failed",
+            ))
+
+    if errors:
+        for e in errors:
+            print(f"Warning: {e}", file=sys.stderr)
+
+    if as_json:
+        import json as _json
+        print(_json.dumps([r.to_dict() for r in reports], indent=2, default=str))
+    else:
+        print(_format_comparison_table(reports, sort_by=sort_by))
+
+
 def cli_help() -> None:
     print("""Solana Rug Guard — On-chain rug-pull detection engine
 
 USAGE:
     python rugguard.py token <MINT_ADDRESS> [--json|--markdown]
     python rugguard.py wallet <WALLET_ADDRESS>
+    python rugguard.py compare <MINT1> <MINT2> [<MINT3> ...] [--json]
     python rugguard.py watch <MINT_ADDRESS> [--interval 60] [--iterations 0]
         [--history PATH] [--webhook URL] [--threshold SCORE]
 
@@ -1673,6 +1853,10 @@ def cli_help() -> None:
     --json        Output as JSON (default for token)
     --markdown    Output as Markdown report
     --md          Alias for --markdown
+    --export csv  Export as CSV (compatible with spreadsheets)
+    --export jsonl
+                  Export as JSONL (one JSON object per line)
+    --sort        Sort tokens in compare output (default: score)
 
 WATCH OPTIONS:
     --interval    Seconds between checks (default: 60)
@@ -1685,6 +1869,8 @@ def cli_help() -> None:
     python rugguard.py token DezXAZ8z7PnrnRJjz3wXBoRgixCa6xjnB7YaB1pPB263
     python rugguard.py token DezXAZ8z7PnrnRJjz3wXBoRgixCa6xjnB7YaB1pPB263 --markdown
     python rugguard.py wallet 9WzDXwBbmkg8ZTbNMqUxvQRAyrZzDsGYdLVL9zYtAWWM
+    python rugguard.py wallet 9WzDXwBbmkg8ZTbNMqUxvQRAyrZzDsGYdLVL9zYtAWWM --export jsonl
+    python rugguard.py compare DezXAZ8z... EPjFWdd5... [--json]
     python rugguard.py watch <MINT_ADDRESS> --iterations 1 --threshold 70
 
 ENVIRONMENT:
@@ -1705,6 +1891,8 @@ def main() -> None:
         cli_token(args)
     elif cmd == "wallet":
         cli_wallet(args)
+    elif cmd == "compare":
+        cli_compare(args)
     elif cmd == "watch":
         cli_watch(args)
     else:
diff --git a/tests/test_checks.py b/tests/test_checks.py
index 27ac3ce..d26fa89 100644
--- a/tests/test_checks.py
+++ b/tests/test_checks.py
@@ -450,4 +450,48 @@ def test_wallet_scan() -> None:
     assert "address" in result
     assert result["address"] == TEST_WALLET
     assert "total_tokens" in result
-    assert isinstance(result["total_tokens"], int)
+    assert isinstance(result['total_tokens'], int)
+
+
+# ── Comparison Tests ──────────────────────────────────────────────────────
+
+class TestCompare:
+    def test_table_two_tokens(self):
+        from rugguard import _format_comparison_table, RugScore, RugFlags, TokenMeta
+        flags = RugFlags()
+        r1 = RugReport(
+            token=TokenMeta(address='A'), safety_score=80, risk_level='LOW',
+            score=RugScore(), flags=flags, warnings=[], recommendation='',
+        )
+        r2 = RugReport(
+            token=TokenMeta(address='B'), safety_score=30, risk_level='HIGH',
+            score=RugScore(), flags=flags, warnings=['Flag'], recommendation='',
+        )
+        tbl = _format_comparison_table([r1, r2])
+        assert 'Safety Score' in tbl
+        assert '80' in tbl
+        assert '30' in tbl
+        assert 'HIGH' in tbl
+        # Rows should be separated by dashes
+        assert '---' in tbl
+
+    def test_sort_by_name(self):
+        from rugguard import _format_comparison_table, RugScore, RugFlags, TokenMeta
+        flags = RugFlags()
+        r1 = RugReport(
+            token=TokenMeta(address='A', symbol='ZZZ'), safety_score=50,
+            risk_level='MEDIUM', score=RugScore(), flags=flags, warnings=[], recommendation='',
+        )
+        r2 = RugReport(
+            token=TokenMeta(address='B', symbol='AAA'), safety_score=50,
+            risk_level='MEDIUM', score=RugScore(), flags=flags, warnings=[], recommendation='',
+        )
+        tbl = _format_comparison_table([r1, r2], sort_by='name')
+        # AAA should come before ZZZ
+        idx_aaa = tbl.index('AAA')
+        idx_zzz = tbl.index('ZZZ')
+        assert idx_aaa < idx_zzz, 'AAA should be before ZZZ when sorted by name'
+
+    def test_empty_no_crash(self):
+        from rugguard import _format_comparison_table
+        assert _format_comparison_table([]) == ''