diff --git a/backend/app/api/connections.py b/backend/app/api/connections.py
index b34423b6..6ef46654 100644
--- a/backend/app/api/connections.py
+++ b/backend/app/api/connections.py
@@ -96,6 +96,8 @@ async def test_connection(
         raise
 
     conn = await session.get(DbConnection, db_connection_uuid)
+    if conn is None:
+        raise HTTPException(status_code=404, detail="connection not found")
     dsn = decrypt_text(conn.dsn_ciphertext, conn.dsn_nonce)
     try:
         version = await probe_database(dsn)
diff --git a/backend/app/api/snapshots.py b/backend/app/api/snapshots.py
index e20467e2..4ffdf778 100644
--- a/backend/app/api/snapshots.py
+++ b/backend/app/api/snapshots.py
@@ -15,6 +15,7 @@
     JobQueue,
     SchemaSnapshot,
     SchemaSnapshotData,
+    TableAnnotation,
 )
 from app.permissions import require_project_member
 from app.schemas import (
@@ -26,6 +27,7 @@
 from app.ddl.export import snapshot_json_to_sql
 from app.ddl.migration import snapshot_diff_to_migration_sql
 from app.diff.schema_diff import diff_snapshots
+from app.spec.data_dictionary import snapshot_to_data_dictionary_md
 from app.jobs.valkey_queue import enqueue_job_signal
 from app.spec.llm import (
     LlmConfigurationError,
@@ -236,6 +238,38 @@ async def export_migration_sql(
     )
 
 
+@router.get(
+    "/{schema_snapshot_uuid}/data-dictionary.md", response_class=PlainTextResponse
+)
+async def export_data_dictionary(
+    schema_snapshot_uuid: uuid.UUID,
+    user: CurrentUser = Depends(get_current_user),
+    session: AsyncSession = Depends(get_read_session),
+) -> str:
+    """Export a snapshot as a Markdown data dictionary, merged with the
+    project's table annotations (living documentation)."""
+    snap = await _get_authorized_snapshot(session, schema_snapshot_uuid, user)
+    if snap is None:
+        return "-- snapshot not found\n"
+    data = await session.get(SchemaSnapshotData, schema_snapshot_uuid)
+    if data is None:
+        return "-- snapshot data not found\n"
+    rows = await session.execute(
+        select(TableAnnotation).where(
+            TableAnnotation.project_space_uuid == snap.project_space_uuid
+        )
+    )
+    annotations = [
+        {
+            "schema_name": a.schema_name,
+            "relation_name": a.relation_name,
+            "body": a.body,
+        }
+        for a in rows.scalars().all()
+    ]
+    return snapshot_to_data_dictionary_md(data.snapshot_json, annotations)
+
+
 @router.get(
     "/{schema_snapshot_uuid}/reversing-spec.md",
     response_class=PlainTextResponse,
diff --git a/backend/app/spec/data_dictionary.py b/backend/app/spec/data_dictionary.py
new file mode 100644
index 00000000..9ac20977
--- /dev/null
+++ b/backend/app/spec/data_dictionary.py
@@ -0,0 +1,176 @@
+"""Render a schema snapshot as a Markdown data dictionary.
+
+Turns a captured snapshot (plus optional project table annotations) into
+"living documentation": one section per table with columns, keys, indexes,
+comments, example values, and any human-authored annotation.
+
+Pure and dialect-agnostic -- it reads the common snapshot JSON shape produced
+by the introspectors, so it works for both PostgreSQL and Snowflake snapshots.
+"""
+
+from __future__ import annotations
+
+from typing import Any, Iterable
+
+
+def _cell(value: object) -> str:
+    """Render a value for a Markdown table cell (escape pipes/newlines)."""
+    if value is None:
+        return ""
+    text = str(value)
+    return text.replace("|", "\\|").replace("\n", " ").strip()
+
+
+def _relation_kind_label(kind: object) -> str:
+    return {
+        "r": "table",
+        "v": "view",
+        "m": "materialized view",
+        "p": "partitioned table",
+        "f": "foreign table",
+    }.get(str(kind or ""), "table")
+
+
+def _annotation_map(
+    annotations: Iterable[dict[str, Any]] | None,
+) -> dict[tuple[str, str], str]:
+    result: dict[tuple[str, str], str] = {}
+    for ann in annotations or []:
+        key = (str(ann.get("schema_name") or ""), str(ann.get("relation_name") or ""))
+        body = ann.get("body")
+        if body:
+            result[key] = str(body)
+    return result
+
+
+def snapshot_to_data_dictionary_md(
+    snapshot: dict[str, Any] | None,
+    annotations: Iterable[dict[str, Any]] | None = None,
+) -> str:
+    """Return a Markdown data dictionary for one snapshot."""
+    snapshot = snapshot or {}
+    relations = snapshot.get("relations") or []
+    columns = snapshot.get("columns") or []
+    pk_columns = snapshot.get("pk_columns") or []
+    fk_edges = snapshot.get("fk_edges") or []
+    indexes = snapshot.get("indexes") or []
+    ann_by_table = _annotation_map(annotations)
+
+    oid_to_rel: dict[Any, dict[str, Any]] = {
+        rel.get("relation_oid"): rel for rel in relations
+    }
+
+    cols_by_oid: dict[Any, list[dict[str, Any]]] = {}
+    for col in columns:
+        cols_by_oid.setdefault(col.get("relation_oid"), []).append(col)
+    for cols in cols_by_oid.values():
+        cols.sort(key=lambda c: (c.get("column_position") or 0))
+
+    pk_by_oid: dict[Any, set[str]] = {}
+    for pk in pk_columns:
+        name = pk.get("column_name")
+        if name is not None:
+            pk_by_oid.setdefault(pk.get("relation_oid"), set()).add(str(name))
+
+    fk_child_cols: dict[Any, set[str]] = {}
+    fk_by_oid: dict[Any, list[dict[str, Any]]] = {}
+    for edge in fk_edges:
+        oid = edge.get("child_relation_oid")
+        fk_by_oid.setdefault(oid, []).append(edge)
+        child_col = edge.get("child_column_name")
+        if child_col is not None:
+            fk_child_cols.setdefault(oid, set()).add(str(child_col))
+
+    idx_by_oid: dict[Any, list[dict[str, Any]]] = {}
+    for idx in indexes:
+        idx_by_oid.setdefault(idx.get("relation_oid"), []).append(idx)
+
+    out: list[str] = ["# Data Dictionary", ""]
+    meta = []
+    if snapshot.get("captured_at"):
+        meta.append(f"Captured: {snapshot['captured_at']}")
+    if snapshot.get("server_version"):
+        meta.append(f"Server: {snapshot['server_version']}")
+    if meta:
+        out.append("_" + " · ".join(meta) + "_")
+        out.append("")
+
+    def _rel_sort_key(rel: dict[str, Any]) -> tuple[str, str]:
+        return (str(rel.get("schema_name") or ""), str(rel.get("relation_name") or ""))
+
+    if not relations:
+        out.append("_No tables in this snapshot._")
+        return "\n".join(out) + "\n"
+
+    for rel in sorted(relations, key=_rel_sort_key):
+        oid = rel.get("relation_oid")
+        schema = str(rel.get("schema_name") or "")
+        name = str(rel.get("relation_name") or "")
+        out.append(f"## {schema}.{name}")
+        kind = _relation_kind_label(rel.get("relation_kind"))
+        if kind != "table":
+            out.append(f"_{kind}_")
+        if rel.get("relation_comment"):
+            out.append(str(rel["relation_comment"]))
+        note = ann_by_table.get((schema, name))
+        if note:
+            out.append(f"> 📝 {note}")
+        out.append("")
+
+        pks = pk_by_oid.get(oid, set())
+        fk_cols = fk_child_cols.get(oid, set())
+        out.append("| # | Column | Type | Null | Default | Key | Comment | Example |")
+        out.append("|---|--------|------|------|---------|-----|---------|---------|")
+        for i, col in enumerate(cols_by_oid.get(oid, []), start=1):
+            col_name = str(col.get("column_name") or "")
+            key_marks = []
+            if col_name in pks:
+                key_marks.append("PK")
+            if col_name in fk_cols:
+                key_marks.append("FK")
+            nullable = "NOT NULL" if col.get("is_not_null") else ""
+            default = col.get("default_expr") if col.get("has_default") else ""
+            out.append(
+                "| {i} | {name} | {type} | {null} | {default} | {key} | {comment} | {example} |".format(
+                    i=i,
+                    name=_cell(col_name),
+                    type=_cell(col.get("data_type")),
+                    null=nullable,
+                    default=_cell(default),
+                    key=" ".join(key_marks),
+                    comment=_cell(col.get("column_comment")),
+                    example=_cell(col.get("example_value")),
+                )
+            )
+        out.append("")
+
+        fks = fk_by_oid.get(oid, [])
+        if fks:
+            out.append("**Foreign keys:**")
+            for edge in fks:
+                parent = oid_to_rel.get(edge.get("parent_relation_oid"), {})
+                parent_name = (
+                    f"{parent.get('schema_name', '')}.{parent.get('relation_name', '')}"
+                    if parent
+                    else "?"
+                )
+                out.append(
+                    f"- `{col_or_q(edge.get('child_column_name'))}` → "
+                    f"`{parent_name}.{col_or_q(edge.get('parent_column_name'))}`"
+                    + (f" ({edge['fk_constraint_name']})" if edge.get("fk_constraint_name") else "")
+                )
+            out.append("")
+
+        idxs = [i for i in idx_by_oid.get(oid, []) if not i.get("is_primary")]
+        if idxs:
+            out.append("**Indexes:**")
+            for idx in idxs:
+                unique = "UNIQUE " if idx.get("is_unique") else ""
+                out.append(f"- {unique}`{idx.get('index_name', '')}`")
+            out.append("")
+
+    return "\n".join(out).rstrip() + "\n"
+
+
+def col_or_q(value: object) -> str:
+    return str(value) if value is not None else "?"
diff --git a/backend/tests/test_data_dictionary.py b/backend/tests/test_data_dictionary.py
new file mode 100644
index 00000000..8d17e603
--- /dev/null
+++ b/backend/tests/test_data_dictionary.py
@@ -0,0 +1,100 @@
+from __future__ import annotations
+
+from app.spec.data_dictionary import snapshot_to_data_dictionary_md
+
+
+def _snapshot():
+    return {
+        "captured_at": "2026-07-05T00:00:00Z",
+        "server_version": "16.2",
+        "relations": [
+            {
+                "relation_oid": 1,
+                "schema_name": "public",
+                "relation_name": "member",
+                "relation_kind": "r",
+                "relation_comment": "회원 마스터",
+            },
+            {
+                "relation_oid": 2,
+                "schema_name": "public",
+                "relation_name": "orders",
+                "relation_kind": "r",
+            },
+            {
+                "relation_oid": 3,
+                "schema_name": "public",
+                "relation_name": "active_members",
+                "relation_kind": "v",
+            },
+        ],
+        "columns": [
+            {"relation_oid": 1, "column_position": 1, "column_name": "member_id", "data_type": "bigint", "is_not_null": True, "example_value": "1001"},
+            {"relation_oid": 1, "column_position": 2, "column_name": "email", "data_type": "varchar(100)", "is_not_null": False, "has_default": True, "default_expr": "''::varchar", "column_comment": "a|b"},
+            {"relation_oid": 2, "column_position": 1, "column_name": "order_id", "data_type": "bigint", "is_not_null": True},
+            {"relation_oid": 2, "column_position": 2, "column_name": "member_id", "data_type": "bigint", "is_not_null": True},
+        ],
+        "pk_columns": [
+            {"relation_oid": 1, "column_name": "member_id", "column_ordinal": 1},
+            {"relation_oid": 2, "column_name": "order_id", "column_ordinal": 1},
+        ],
+        "fk_edges": [
+            {
+                "fk_constraint_name": "fk_orders_member",
+                "child_relation_oid": 2,
+                "parent_relation_oid": 1,
+                "child_column_name": "member_id",
+                "parent_column_name": "member_id",
+                "column_ordinal": 1,
+            }
+        ],
+        "indexes": [
+            {"relation_oid": 1, "index_name": "member_pkey", "is_unique": True, "is_primary": True},
+            {"relation_oid": 1, "index_name": "ix_member_email", "is_unique": True, "is_primary": False},
+        ],
+    }
+
+
+def test_renders_tables_columns_keys_and_metadata():
+    md = snapshot_to_data_dictionary_md(_snapshot())
+    assert md.startswith("# Data Dictionary")
+    assert "Captured: 2026-07-05T00:00:00Z" in md
+    assert "## public.member" in md
+    assert "회원 마스터" in md
+    # column row with PK marker + example
+    assert "| member_id | bigint | NOT NULL |" in md
+    assert "| PK |" in md or " PK " in md
+    assert "1001" in md
+    # default rendered when has_default
+    assert "''::varchar" in md
+    # pipe in comment escaped
+    assert "a\\|b" in md
+
+
+def test_renders_foreign_keys_and_non_primary_indexes():
+    md = snapshot_to_data_dictionary_md(_snapshot())
+    assert "**Foreign keys:**" in md
+    assert "`member_id` → `public.member.member_id`" in md
+    assert "(fk_orders_member)" in md
+    assert "**Indexes:**" in md
+    assert "UNIQUE `ix_member_email`" in md
+    # primary-key index is not listed under Indexes
+    assert "member_pkey" not in md.split("**Indexes:**", 1)[-1]
+
+
+def test_merges_project_annotations():
+    md = snapshot_to_data_dictionary_md(
+        _snapshot(),
+        annotations=[
+            {"schema_name": "public", "relation_name": "orders", "body": "주문 트랜잭션 테이블"}
+        ],
+    )
+    assert "> 📝 주문 트랜잭션 테이블" in md
+
+
+def test_labels_views_and_handles_empty_snapshot():
+    md = snapshot_to_data_dictionary_md(_snapshot())
+    assert "## public.active_members" in md
+    assert "_view_" in md
+    empty = snapshot_to_data_dictionary_md({})
+    assert "No tables in this snapshot" in empty
diff --git a/frontend/src/App.tsx b/frontend/src/App.tsx
index 6d8675ce..6d7258b4 100644
--- a/frontend/src/App.tsx
+++ b/frontend/src/App.tsx
@@ -36,6 +36,7 @@ import {
   deleteAnnotation,
   deleteView,
   diffSnapshots,
+  fetchDataDictionaryMarkdown,
   getSnapshot,
   getView,
   listAnnotations,
@@ -805,6 +806,16 @@ export default function App() {
     downloadText("pg-erd-diagram.mermaid", exportMermaid(nodes, edges), "text/plain");
   }
 
+  async function onDownloadDataDictionary() {
+    if (!snapshotId) return;
+    try {
+      const md = await fetchDataDictionaryMarkdown(snapshotId);
+      downloadText("data-dictionary.md", md, "text/markdown");
+    } catch {
+      window.alert("데이터 사전을 불러오지 못했습니다.");
+    }
+  }
+
   function onRelDelete() {
     if (!editingEdge) return;
     if (!window.confirm("정말로 이 관계를 삭제하시겠습니까?")) return;
@@ -1699,6 +1710,19 @@ export default function App() {
             >
               {"{}"}
             </button>
+            <button
+              type="button"
+              onClick={onDownloadDataDictionary}
+              disabled={!snapshotId}
+              title={
+                !snapshotId
+                  ? "스냅샷을 먼저 생성하세요"
+                  : "데이터 사전 (Markdown) 내보내기"
+              }
+              aria-label="데이터 사전 내보내기"
+            >
+              📖
+            </button>
             <div className="srOnly" aria-live="polite">
               {[layoutMessage, nodeSearchStatus].filter(Boolean).join(" ")}
             </div>
diff --git a/frontend/src/api.ts b/frontend/src/api.ts
index 0cbca918..c64f41a5 100644
--- a/frontend/src/api.ts
+++ b/frontend/src/api.ts
@@ -401,6 +401,18 @@ export async function deleteView(viewId: string): Promise<void> {
   if (!r.ok) throw new Error(`deleteView failed: ${r.status}`)
 }
 
+export async function fetchDataDictionaryMarkdown(snapshotId: string): Promise<string> {
+  if (DEMO_MODE) {
+    return '# Data Dictionary (demo)\n\n## public.member\n회원 마스터\n\n| # | Column | Type | Null | Key |\n|---|--------|------|------|-----|\n| 1 | member_id | bigint | NOT NULL | PK |\n'
+  }
+  const r = await fetch(
+    `${API_BASE}/api/snapshots/${encodeURIComponent(snapshotId)}/data-dictionary.md`,
+    { credentials: 'include' }
+  )
+  if (!r.ok) throw new Error(`fetchDataDictionaryMarkdown failed: ${r.status}`)
+  return r.text()
+}
+
 export async function testConnection(connectionId: string): Promise<ConnectionTestResult> {
   if (DEMO_MODE) {
     return { ok: true, server_version: 'PostgreSQL 16.2 (demo)', error: null }