diff --git a/Cargo.lock b/Cargo.lock
index 75a9dbae4e5e..e5fe7cbc3bfb 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -4531,6 +4531,16 @@ dependencies = [
  "trie-rs",
 ]
 
+[[package]]
+name = "tabby-repositories"
+version = "0.10.0-dev.0"
+dependencies = [
+ "anyhow",
+ "serde_json",
+ "sqlx",
+ "tabby-common",
+]
+
 [[package]]
 name = "tabby-scheduler"
 version = "0.10.0-dev.0"
diff --git a/Cargo.toml b/Cargo.toml
index 7712878600ec..f385bb6159b8 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -6,12 +6,13 @@ members = [
     "crates/tabby-scheduler",
     "crates/tabby-download",
     "crates/tabby-inference",
+    "crates/tabby-repositories",
     "crates/llama-cpp-bindings",
     "crates/http-api-bindings",
     "crates/aim-downloader",
     "crates/juniper-axum",
     "ee/tabby-webserver",
-    "ee/tabby-db", "ee/tabby-db-macros",
+    "ee/tabby-db", "ee/tabby-db-macros", "crates/tabby-repositories",
 ]
 
 [workspace.package]
@@ -21,6 +22,7 @@ authors = ["Meng Zhang"]
 homepage = "https://github.com/TabbyML/tabby"
 
 [workspace.dependencies]
+sqlx = { version = "0.7.3", features = ["sqlite", "chrono", "runtime-tokio", "macros"] }
 lazy_static = "1.4.0"
 serde = { version = "1.0", features = ["derive"] }
 serde_json = "1"
diff --git a/crates/tabby-common/src/path.rs b/crates/tabby-common/src/path.rs
index 823d64f91ce4..25b0f491ca35 100644
--- a/crates/tabby-common/src/path.rs
+++ b/crates/tabby-common/src/path.rs
@@ -29,6 +29,10 @@ pub fn config_file() -> PathBuf {
     tabby_root().join("config.toml")
 }
 
+pub fn repository_meta_db() -> PathBuf {
+    tabby_root().join("repositories.sqlite")
+}
+
 pub fn usage_id_file() -> PathBuf {
     tabby_root().join("usage_anonymous_id")
 }
diff --git a/crates/tabby-repositories/.env b/crates/tabby-repositories/.env
new file mode 100644
index 000000000000..5f4d57ee15eb
--- /dev/null
+++ b/crates/tabby-repositories/.env
@@ -0,0 +1 @@
+DATABASE_URL=sqlite://crates/tabby-repositories/schema.sqlite
diff --git a/crates/tabby-repositories/Cargo.toml b/crates/tabby-repositories/Cargo.toml
new file mode 100644
index 000000000000..a5960dac3577
--- /dev/null
+++ b/crates/tabby-repositories/Cargo.toml
@@ -0,0 +1,14 @@
+[package]
+name = "tabby-repositories"
+version.workspace = true
+edition.workspace = true
+authors.workspace = true
+homepage.workspace = true
+
+# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
+
+[dependencies]
+sqlx = { workspace = true }
+anyhow = { workspace = true }
+tabby-common = { path = "../tabby-common" }
+serde_json = { workspace = true }
diff --git a/crates/tabby-repositories/schema.sql b/crates/tabby-repositories/schema.sql
new file mode 100644
index 000000000000..c98c37dd85be
--- /dev/null
+++ b/crates/tabby-repositories/schema.sql
@@ -0,0 +1,10 @@
+DROP TABLE IF EXISTS repository_meta;
+CREATE TABLE repository_meta(
+    git_url TEXT NOT NULL,
+    filepath TEXT NOT NULL,
+    language TEXT NOT NULL,
+    max_line_length INTEGER NOT NULL,
+    avg_line_length REAL NOT NULL,
+    alphanum_fraction REAL NOT NULL,
+    tags TEXT NOT NULL
+);
diff --git a/crates/tabby-repositories/schema.sqlite b/crates/tabby-repositories/schema.sqlite
new file mode 100644
index 000000000000..8ee377f05eb2
Binary files /dev/null and b/crates/tabby-repositories/schema.sqlite differ
diff --git a/crates/tabby-repositories/src/lib.rs b/crates/tabby-repositories/src/lib.rs
new file mode 100644
index 000000000000..01ea76052559
--- /dev/null
+++ b/crates/tabby-repositories/src/lib.rs
@@ -0,0 +1,57 @@
+use anyhow::Result;
+use sqlx::sqlite::SqlitePoolOptions;
+use sqlx::{query, SqlitePool};
+use sqlx::{sqlite::SqliteConnectOptions, Pool, Sqlite};
+use std::str::FromStr;
+use tabby_common::Tag;
+
+pub struct RepositoryCache {
+    pool: Pool<Sqlite>,
+}
+
+struct RepositoryMetaDAO {
+    git_url: String,
+    filepath: String,
+    language: String,
+    max_line_length: usize,
+    avg_line_length: f32,
+    alphanum_fraction: f32,
+    tags: String,
+}
+
+impl RepositoryCache {
+    pub async fn new() -> Result<Self> {
+        let init_query = include_str!("../schema.sql");
+        let options = SqliteConnectOptions::new()
+            .filename(tabby_common::path::repository_meta_db())
+            .create_if_missing(true);
+        let pool = SqlitePool::connect_with(options).await?;
+        sqlx::query(init_query).execute(&pool).await?;
+        Ok(RepositoryCache { pool })
+    }
+
+    pub async fn clear(&self) -> Result<()> {
+        query!("DELETE FROM repository_meta")
+            .execute(&self.pool)
+            .await?;
+        Ok(())
+    }
+
+    pub async fn add_repository_meta(
+        &self,
+        git_url: String,
+        filepath: String,
+        language: String,
+        max_line_length: i64,
+        avg_line_length: f32,
+        alphanum_fraction: f32,
+        tags: Vec<Tag>,
+    ) -> Result<()> {
+        let tags = serde_json::to_string(&tags)?;
+        query!("INSERT INTO repository_meta (git_url, filepath, language, max_line_length, avg_line_length, alphanum_fraction, tags)
+                VALUES ($1, $2, $3, $4, $5, $6, $7)",
+                git_url, filepath, language, max_line_length, avg_line_length, alphanum_fraction, tags
+        ).execute(&self.pool).await?;
+        Ok(())
+    }
+}
diff --git a/ee/tabby-db/Cargo.toml b/ee/tabby-db/Cargo.toml
index 9f8f288068f7..e6c584281bbe 100644
--- a/ee/tabby-db/Cargo.toml
+++ b/ee/tabby-db/Cargo.toml
@@ -15,8 +15,8 @@ anyhow.workspace = true
 chrono = { workspace = true, features = ["serde"] }
 hash-ids.workspace = true
 lazy_static.workspace = true
+sqlx = { workspace = true }
 sql_query_builder = { version = "2.1.0", features = ["sqlite"] }
-sqlx = { version = "0.7.3", features = ["sqlite", "chrono", "runtime-tokio", "macros"] }
 tabby-common = { path = "../../crates/tabby-common" }
 tokio = { workspace = true, features = ["fs"] }
 uuid.workspace = true