feat(usage): include Claude sub-agent usage in statistics

src-tauri/src/services/session_usage.rs

@@ -1,11 +1,12 @@
 //! Claude Code 会话日志使用追踪
 //!
 //! 从 ~/.claude/projects/ 下的 JSONL 会话文件中提取 token 使用数据，
-//! 实现无代理模式下的使用统计。
+//! 实现无代理模式下的使用统计。同时解析子 Agent (sub-agent) 的对话记录。
 //!
 //! ## 数据流
 //! ```text
-//! ~/.claude/projects/*/*.jsonl → 增量解析 → 去重 → 费用计算 → proxy_request_logs 表
+//! ~/.claude/projects/*/*.jsonl                → 增量解析 → 去重 → 费用计算 → proxy_request_logs 表
+//! ~/.claude/projects/*/<sessionId>/subagents/ → 增量解析 → 去重 → 费用计算 → proxy_request_logs 表
 //! ```
 
 use crate::config::get_claude_config_dir;
@@ -54,7 +55,7 @@ struct ParsedAssistantUsage {
     session_id: Option<String>,
 }
 
-/// 同步 Claude Code 会话日志到使用统计数据库
+/// 同步 Claude Code 会话日志到使用统计数据库（包括子 Agent 记录）
 pub fn sync_claude_session_logs(db: &Database) -> Result<SessionSyncResult, AppError> {
     let projects_dir = get_claude_config_dir().join("projects");
     if !projects_dir.exists() {
@@ -73,13 +74,13 @@ pub fn sync_claude_session_logs(db: &Database) -> Result<SessionSyncResult, AppE
         errors: vec![],
     };
 
-    // 收集所有 .jsonl 文件
+    // 收集主会话 .jsonl 文件
     let jsonl_files = collect_jsonl_files(&projects_dir);
 
     for file_path in &jsonl_files {
         result.files_scanned += 1;
 
-        match sync_single_file(db, file_path) {
+        match sync_single_file(db, file_path, "session_log", "session:") {
             Ok((imported, skipped)) => {
                 result.imported += imported;
                 result.skipped += skipped;
@@ -92,12 +93,32 @@ pub fn sync_claude_session_logs(db: &Database) -> Result<SessionSyncResult, AppE
         }
     }
 
+    // 收集子 Agent 会话文件
+    let subagent_files = collect_subagent_files(&projects_dir);
+
+    for file_path in &subagent_files {
+        result.files_scanned += 1;
+
+        match sync_single_file(db, file_path, "session_subagent", "subagent:") {
+            Ok((imported, skipped)) => {
+                result.imported += imported;
+                result.skipped += skipped;
+            }
+            Err(e) => {
+                let msg = format!("{}: {e}", file_path.display());
+                log::warn!("[SESSION-SYNC] 子Agent文件解析失败: {msg}");
+                result.errors.push(msg);
+            }
+        }
+    }
+
     if result.imported > 0 {
         log::info!(
-            "[SESSION-SYNC] 同步完成: 导入 {} 条, 跳过 {} 条, 扫描 {} 个文件",
+            "[SESSION-SYNC] 同步完成: 导入 {} 条, 跳过 {} 条, 扫描 {} 个文件 (含 {} 个子Agent文件)",
             result.imported,
             result.skipped,
-            result.files_scanned
+            result.files_scanned,
+            subagent_files.len()
         );
     }
 
@@ -132,8 +153,87 @@ fn collect_jsonl_files(projects_dir: &Path) -> Vec<PathBuf> {
     files
 }
 
+/// 收集所有子 Agent 的会话文件
+///
+/// 扫描路径: ~/.claude/projects/<project>/<sessionId>/subagents/**/agent-*.jsonl
+fn collect_subagent_files(projects_dir: &Path) -> Vec<PathBuf> {
+    let mut files = Vec::new();
+
+    let entries = match fs::read_dir(projects_dir) {
+        Ok(e) => e,
+        Err(_) => return files,
+    };
+
+    for entry in entries.flatten() {
+        let project_dir = entry.path();
+        if !project_dir.is_dir() {
+            continue;
+        }
+
+        // 遍历项目目录下的每个子目录 (可能是 session 目录)
+        let session_entries = match fs::read_dir(&project_dir) {
+            Ok(e) => e,
+            Err(_) => continue,
+        };
+
+        for session_entry in session_entries.flatten() {
+            let session_path = session_entry.path();
+            if !session_path.is_dir() {
+                continue;
+            }
+
+            let subagents_dir = session_path.join("subagents");
+            if !subagents_dir.is_dir() {
+                continue;
+            }
+
+            collect_subagent_files_recursive(&subagents_dir, &mut files);
+        }
+    }
+
+    files
+}
+
+/// 递归扫描 subagents 目录下任意层级的 agent-*.jsonl。
+///
+/// Claude Code/工作流可能把子 Agent 日志放在 subagents/workflows/<runId>/ 等更深层目录，
+/// 因此这里不限制递归深度，只在文件名层面过滤实际需要导入的 transcript。
+fn collect_subagent_files_recursive(dir: &Path, files: &mut Vec<PathBuf>) {
+    let entries = match fs::read_dir(dir) {
+        Ok(e) => e,
+        Err(_) => return,
+    };
+
+    for entry in entries.flatten() {
+        let file_type = match entry.file_type() {
+            Ok(ft) => ft,
+            Err(_) => continue,
+        };
+        let path = entry.path();
+
+        if file_type.is_dir() {
+            collect_subagent_files_recursive(&path, files);
+        } else if file_type.is_file() && is_subagent_jsonl_file(&path) {
+            files.push(path);
+        }
+    }
+}
+
+fn is_subagent_jsonl_file(path: &Path) -> bool {
+    let file_name = path.file_name().and_then(|n| n.to_str()).unwrap_or("");
+    file_name.starts_with("agent-") && path.extension().and_then(|e| e.to_str()) == Some("jsonl")
+}
+
 /// 同步单个 JSONL 文件，返回 (imported, skipped)
-fn sync_single_file(db: &Database, file_path: &Path) -> Result<(u32, u32), AppError> {
+///
+/// `data_source` 用于区分数据来源 ("session_log" 或 "session_subagent")
+/// `request_id_prefix` 用于生成唯一的 request_id 前缀
+fn sync_single_file(
+    db: &Database,
+    file_path: &Path,
+    data_source: &str,
+    request_id_prefix: &str,
+) -> Result<(u32, u32), AppError> {
     let file_path_str = file_path.to_string_lossy().to_string();
 
     // 获取文件元数据
@@ -278,14 +378,14 @@ fn sync_single_file(db: &Database, file_path: &Path) -> Result<(u32, u32), AppEr
             continue;
         }
 
-        let request_id = format!("session:{}", msg.message_id);
+        let request_id = format!("{}{}", request_id_prefix, msg.message_id);
 
         // 跳过 output_tokens 为 0 的无意义条目
         if msg.output_tokens == 0 {
             continue;
         }
 
-        match insert_session_log_entry(db, &request_id, msg) {
+        match insert_session_log_entry(db, &request_id, msg, data_source) {
             Ok(true) => imported += 1,
             Ok(false) => skipped += 1,
             Err(e) => {
@@ -335,10 +435,13 @@ fn update_sync_state(
 }
 
 /// 插入单条会话日志到 proxy_request_logs，返回是否成功插入 (true=新插入, false=已存在)
+///
+/// `data_source` 区分主会话 ("session_log") 和子 Agent ("session_subagent")
 fn insert_session_log_entry(
     db: &Database,
     request_id: &str,
     msg: &ParsedAssistantUsage,
+    data_source: &str,
 ) -> Result<bool, AppError> {
     let conn = lock_conn!(db.conn);
 
@@ -432,11 +535,11 @@ fn insert_session_log_entry(
             200i64,             // status_code: 有 stop_reason 说明请求成功
             Option::<String>::None, // error_message
             msg.session_id,
-            Some("session_log"), // provider_type
+            Some(data_source),  // provider_type: 使用 data_source 值
             1i64,               // is_streaming: Claude Code 通常使用流式
             "1.0",              // cost_multiplier
             created_at,
-            "session_log",      // data_source
+            data_source,        // data_source: 区分主会话和子Agent
         ],
     )
     .map_err(|e| AppError::Database(format!("插入会话日志失败: {e}")))?;
@@ -527,7 +630,11 @@ pub fn get_data_source_breakdown(db: &Database) -> Result<Vec<DataSourceSummary>
     let conn = lock_conn!(db.conn);
 
     let mut stmt = conn.prepare(
-        "SELECT COALESCE(data_source, 'proxy') as ds, COUNT(*) as cnt,
+        "SELECT CASE COALESCE(data_source, 'proxy')
+                    WHEN 'session_subagent' THEN 'session_log'
+                    ELSE COALESCE(data_source, 'proxy')
+                END as ds,
+                COUNT(*) as cnt,
                 COALESCE(SUM(CAST(total_cost_usd AS REAL)), 0) as cost
          FROM proxy_request_logs
          GROUP BY ds
@@ -553,6 +660,7 @@ pub fn get_data_source_breakdown(db: &Database) -> Result<Vec<DataSourceSummary>
 #[cfg(test)]
 mod tests {
     use super::*;
+    use std::fs;
 
     #[test]
     fn test_parse_usage_from_jsonl_line() {
@@ -631,4 +739,43 @@ mod tests {
         messages.insert("msg_1".to_string(), final_entry);
         assert_eq!(messages.get("msg_1").unwrap().output_tokens, 1349);
     }
+
+    #[test]
+    fn test_collect_subagent_files_recurses_nested_workflow_dirs() {
+        let temp = tempfile::tempdir().unwrap();
+        let subagents_dir = temp
+            .path()
+            .join("project-a")
+            .join("session-1")
+            .join("subagents");
+        let nested_dir = subagents_dir.join("workflows").join("run-1");
+
+        fs::create_dir_all(&nested_dir).unwrap();
+        fs::write(subagents_dir.join("agent-root.jsonl"), "").unwrap();
+        fs::write(nested_dir.join("agent-nested.jsonl"), "").unwrap();
+        fs::write(nested_dir.join("not-agent.jsonl"), "").unwrap();
+        fs::write(subagents_dir.join("agent-not-json.txt"), "").unwrap();
+
+        let mut files: Vec<PathBuf> = collect_subagent_files(temp.path())
+            .into_iter()
+            .map(|path| path.strip_prefix(temp.path()).unwrap().to_path_buf())
+            .collect();
+        files.sort();
+
+        let mut expected = vec![
+            PathBuf::from("project-a")
+                .join("session-1")
+                .join("subagents")
+                .join("agent-root.jsonl"),
+            PathBuf::from("project-a")
+                .join("session-1")
+                .join("subagents")
+                .join("workflows")
+                .join("run-1")
+                .join("agent-nested.jsonl"),
+        ];
+        expected.sort();
+
+        assert_eq!(files, expected);
+    }
 }

src-tauri/src/services/usage_stats.rs

@@ -120,6 +120,9 @@ pub struct RequestLogDetail {
 /// SQL fragment: resolve provider_name with fallback for session-based entries.
 /// Session logs use placeholder provider_ids (_session, _codex_session, _gemini_session)
 /// that don't exist in the providers table — this COALESCE gives them readable names.
+///
+/// This helper is shared by raw log and daily rollup queries. Keep it independent from
+/// proxy_request_logs-only columns such as data_source; usage_daily_rollups does not store them.
 fn provider_name_coalesce(log_alias: &str, provider_alias: &str) -> String {
     format!(
         "COALESCE({provider_alias}.name, CASE {log_alias}.provider_id \
@@ -130,6 +133,16 @@ fn provider_name_coalesce(log_alias: &str, provider_alias: &str) -> String {
     )
 }
 
+/// UI 展示层不单独拆分 Claude 主会话和子 Agent，会统一显示为 session_log。
+/// 数据库仍保留原始 data_source，便于后续排查或重新拆分。
+fn display_data_source_expr(log_alias: &str) -> String {
+    format!(
+        "CASE COALESCE({log_alias}.data_source, 'proxy') \
+         WHEN 'session_subagent' THEN 'session_log' \
+         ELSE COALESCE({log_alias}.data_source, 'proxy') END"
+    )
+}
+
 impl Database {
     /// 获取使用量汇总
     pub fn get_usage_summary(
@@ -671,13 +684,14 @@ impl Database {
         params.push(Box::new(offset as i64));
 
         let logs_pname = provider_name_coalesce("l", "p");
+        let logs_data_source = display_data_source_expr("l");
         let sql = format!(
             "SELECT l.request_id, l.provider_id, {logs_pname} as provider_name, l.app_type, l.model,
                     l.request_model, l.cost_multiplier,
                     l.input_tokens, l.output_tokens, l.cache_read_tokens, l.cache_creation_tokens,
                     l.input_cost_usd, l.output_cost_usd, l.cache_read_cost_usd, l.cache_creation_cost_usd, l.total_cost_usd,
                     l.is_streaming, l.latency_ms, l.first_token_ms, l.duration_ms,
-                    l.status_code, l.error_message, l.created_at, l.data_source
+                    l.status_code, l.error_message, l.created_at, {logs_data_source} as data_source
              FROM proxy_request_logs l
              LEFT JOIN providers p ON l.provider_id = p.id AND l.app_type = p.app_type
              {where_clause}
@@ -749,13 +763,14 @@ impl Database {
         let conn = lock_conn!(self.conn);
 
         let detail_pname = provider_name_coalesce("l", "p");
+        let detail_data_source = display_data_source_expr("l");
         let detail_sql = format!(
             "SELECT l.request_id, l.provider_id, {detail_pname} as provider_name, l.app_type, l.model,
                     l.request_model, l.cost_multiplier,
                     input_tokens, output_tokens, cache_read_tokens, cache_creation_tokens,
                     input_cost_usd, output_cost_usd, cache_read_cost_usd, cache_creation_cost_usd, total_cost_usd,
                     is_streaming, latency_ms, first_token_ms, duration_ms,
-                    status_code, error_message, created_at, l.data_source
+                    status_code, error_message, created_at, {detail_data_source} as data_source
              FROM proxy_request_logs l
              LEFT JOIN providers p ON l.provider_id = p.id AND l.app_type = p.app_type
              WHERE l.request_id = ?"