reorg skill folder

backend/app/agent/chart.py

@@ -44,7 +44,7 @@ CHART_INSTRUCTIONS = """
     - For daily question, the time unit should be "yearmonthdate".
     - Default time unit is "yearmonth".
 - For each axis, generate the corresponding human-readable title based on the language provided by the user.
-- Make sure all of the fields(x, y, xOffset, color, etc.) in the encoding section of the chart schema are present in the column names of the data.
+- **CRITICAL REQUIREMENT**: Make sure all of the `field` values in the encoding section of the chart schema EXACTLY MATCH the column names of the sample data provided! DO NOT translate, rename, or hallucinate `field` names. If you want to show a translated name in the chart, use the `title` property, NOT the `field` property!
 
 ### GUIDELINES TO PLOT CHART ###
 
@@ -233,6 +233,18 @@ Language: Chinese (Simplified)
             
         content = content.strip()
         result = json.loads(content)
+        
+        # Post-process to fix common LLM hallucinations (translating field names)
+        if result.get("chart_spec") and isinstance(result["chart_spec"], dict):
+            encoding = result["chart_spec"].get("encoding", {})
+            for channel, enc_def in encoding.items():
+                if isinstance(enc_def, dict) and "field" in enc_def:
+                    field = enc_def["field"]
+                    # If field is not in columns, try to find a match or let it be (Vega will render empty)
+                    # But if we can detect it was translated, we might not be able to fix it perfectly.
+                    # As a simple fallback, if there's only one string column and one numeric column, we could guess,
+                    # but it's safer to just rely on the stricter prompt.
+        
         return ChartGenerationResponse(**result)
         
     except Exception as e:

backend/app/agent/nl2sql.py

@@ -4,7 +4,6 @@ import os
 import json
 import time
 import threading
-import re
 from pathlib import Path
 from typing import List, Optional, Dict, Any, Callable, Awaitable
 from pydantic import BaseModel, Field
@@ -35,9 +34,11 @@ MAX_UPLOAD_CACHE_ITEMS = 8
 NL2SQL_MAX_TOKENS = 900
 NL2SQL_TEMPERATURE = 0.1
 NL2SQL_REASONING_EFFORT = "low"
-NL2SQL_LLM_TIMEOUT_SECONDS = 60*5
+NL2SQL_LLM_TIMEOUT_SECONDS = int(os.getenv("NL2SQL_LLM_TIMEOUT_SECONDS", "90"))
+NL2SQL_LLM_REQUEST_TIMEOUT_SECONDS = int(os.getenv("NL2SQL_LLM_REQUEST_TIMEOUT_SECONDS", "45"))
+NL2SQL_LLM_RETRY_COUNT = int(os.getenv("NL2SQL_LLM_RETRY_COUNT", "0"))
 NL2SQL_SQL_EXEC_TIMEOUT_SECONDS = 60
-NL2SQL_CHART_TIMEOUT_SECONDS = 45
+NL2SQL_CHART_TIMEOUT_SECONDS = int(os.getenv("NL2SQL_CHART_TIMEOUT_SECONDS", "20"))
 
 _schema_cache: Dict[str, Dict[str, Any]] = {}
 _connection_cache: Dict[str, Dict[str, Any]] = {}
@@ -163,6 +164,23 @@ def _execute_upload_sql(sql_query: str, df: pd.DataFrame) -> List[Dict[str, Any]
     conn.close()
     return result_df.to_dict(orient="records")
 
+def _to_number(value: Any) -> Optional[float]:
+    if isinstance(value, bool):
+        return None
+    if isinstance(value, (int, float)):
+        return float(value)
+    if isinstance(value, str):
+        text = value.strip().replace(",", "")
+        if not text:
+            return None
+        try:
+            return float(text)
+        except ValueError:
+            return None
+    return None
+
+# _build_fallback_chart removed as per user request to not hardcode fallbacks
+
 def _build_schema_cache_key(source: str, connector: Any) -> str:
     # If source is ds:ID, that's already a good key
     if source.startswith("ds:"):
@@ -270,7 +288,7 @@ async def process_nl2sql(
     if connector:
         await emit_progress("正在检测数据源连通性")
         cached_schema = _get_cached_schema(request.source, connector)
-        if cached_schema:
+        if cached_schema is not None:
             schema = cached_schema
             await emit_progress(f"命中 Schema 缓存，已加载 {len(schema)} 张表")
         else:
@@ -290,23 +308,6 @@ async def process_nl2sql(
             _set_cached_schema(request.source, connector, schema)
             await emit_progress(f"Schema 拉取完成，共 {len(schema)} 张表 ({time.perf_counter() - schema_started:.2f}s)")
          
-    if connector and not schema:
-         retry_started = time.perf_counter()
-         # Double check in case schema was empty but connection is ok (e.g. empty db)
-         if not await _check_connection_with_cache(request.source, connector):
-              return NL2SQLResponse(sql="", result=[], error=f"Failed to connect to {request.source}")
-         
-         try:
-             schema = await asyncio.wait_for(
-                 asyncio.to_thread(connector.get_schema),
-                 timeout=30.0
-             )
-         except Exception as e:
-             return NL2SQLResponse(sql="", result=[], error=f"Failed to fetch schema on retry: {e}")
-             
-         _set_cached_schema(request.source, connector, schema)
-         await emit_progress(f"Schema 二次拉取完成，共 {len(schema)} 张表 ({time.perf_counter() - retry_started:.2f}s)")
-         
     schema_str = json.dumps(schema, ensure_ascii=False, separators=(",", ":"))
 
     # Try to load MDL context
@@ -378,18 +379,45 @@ Language: Chinese (Simplified)
     try:
         llm_started = time.perf_counter()
         await emit_progress("正在生成 SQL")
+        response = None
+        last_error = ""
+
+        for attempt in range(NL2SQL_LLM_RETRY_COUNT + 1):
+            try:
                 response = await asyncio.wait_for(
                     provider.chat(
                         messages=messages,
                         max_tokens=NL2SQL_MAX_TOKENS,
                         temperature=NL2SQL_TEMPERATURE,
                         reasoning_effort=NL2SQL_REASONING_EFFORT,
+                        request_timeout=NL2SQL_LLM_REQUEST_TIMEOUT_SECONDS,
+                        num_retries=0,
                     ),
                     timeout=NL2SQL_LLM_TIMEOUT_SECONDS,
                 )
+            except asyncio.TimeoutError:
+                last_error = f"LLM generation timeout after {NL2SQL_LLM_TIMEOUT_SECONDS}s"
+                if attempt < NL2SQL_LLM_RETRY_COUNT:
+                    await emit_progress(f"SQL 生成超时，正在重试 ({attempt + 1}/{NL2SQL_LLM_RETRY_COUNT})")
+                    continue
+                return NL2SQLResponse(sql="", result=[], error=last_error)
+            except Exception as e:
+                last_error = f"LLM generation failed: {e}"
+                if attempt < NL2SQL_LLM_RETRY_COUNT:
+                    await emit_progress(f"SQL 生成失败，正在重试 ({attempt + 1}/{NL2SQL_LLM_RETRY_COUNT})")
+                    continue
+                return NL2SQLResponse(sql="", result=[], error=last_error)
 
             if response.finish_reason == "error":
-            return NL2SQLResponse(sql="", result=[], error=response.content or "LLM Error")
+                last_error = response.content or "LLM Error"
+                if attempt < NL2SQL_LLM_RETRY_COUNT:
+                    await emit_progress(f"模型返回错误，正在重试 ({attempt + 1}/{NL2SQL_LLM_RETRY_COUNT})")
+                    continue
+                return NL2SQLResponse(sql="", result=[], error=last_error)
+            break
+
+        if response is None:
+            return NL2SQLResponse(sql="", result=[], error=last_error or "LLM generation failed")
 
         content = (response.content or "").strip()
         if not content:
@@ -406,14 +434,11 @@ Language: Chinese (Simplified)
         try:
             result_json = json.loads(content)
             sql_query = result_json.get("sql", "").strip()
-            reasoning = result_json.get("reasoning", "") # We can log this or return it if needed
         except json.JSONDecodeError:
             # Fallback if LLM doesn't return valid JSON despite instructions
             sql_query = content
         await emit_progress(f"SQL 生成完成 ({time.perf_counter() - llm_started:.2f}s)")
 
-    except asyncio.TimeoutError:
-        return NL2SQLResponse(sql="", result=[], error=f"LLM generation timeout after {NL2SQL_LLM_TIMEOUT_SECONDS}s")
     except Exception as e:
         return NL2SQLResponse(sql="", result=[], error=f"LLM generation failed: {e}")
 
@@ -481,13 +506,23 @@ Language: Chinese (Simplified)
                 generate_chart(formatted_results, request.query),
                 timeout=NL2SQL_CHART_TIMEOUT_SECONDS,
             )
+            if not chart_response or not chart_response.chart_spec:
+                # Do not fallback automatically if the LLM explicitly decided not to or failed.
+                # Just pass whatever it returned (or lack thereof)
+                pass
             await emit_progress(f"可视化方案生成完成 ({time.perf_counter() - chart_started:.2f}s)")
         await emit_progress(f"NL2SQL 总耗时 {time.perf_counter() - total_started:.2f}s")
 
         return NL2SQLResponse(sql=sql_query, result=formatted_results, chart=chart_response)
     except asyncio.TimeoutError:
         if timeout_stage == "chart_generation":
-            return NL2SQLResponse(sql=sql_query, result=formatted_results, error=f"Chart generation timeout after {NL2SQL_CHART_TIMEOUT_SECONDS}s")
+            fallback_chart = ChartGenerationResponse(
+                reasoning=f"Chart generation timeout after {NL2SQL_CHART_TIMEOUT_SECONDS}s",
+                chart_type="",
+                can_visualize=False,
+                chart_spec=None,
+            )
+            return NL2SQLResponse(sql=sql_query, result=formatted_results, chart=fallback_chart)
         return NL2SQLResponse(sql=sql_query, result=[], error=f"SQL execution timeout after {NL2SQL_SQL_EXEC_TIMEOUT_SECONDS}s")
     except Exception as e:
         return NL2SQLResponse(sql=sql_query, result=[], error=f"SQL execution failed: {e}")

backend/app/api/skills.py

@@ -3,6 +3,7 @@ import os
 import shutil
 import zipfile
 import tarfile
+import re
 import yaml
 from typing import List, Optional, Dict, Any
 from datetime import datetime
@@ -111,6 +112,26 @@ def _save_data(data: List[Dict[str, Any]]):
     with open(DATA_FILE, "w") as f:
         json.dump(data, f, indent=2, ensure_ascii=False)
 
+def _safe_skill_dir_name(value: str) -> str:
+    safe = re.sub(r'[^a-zA-Z0-9_\-]', '_', value or "").lower()
+    return safe or "skill"
+
+def _write_skill_markdown(skill_dir: str, skill_name: str, description: Optional[str], content: str) -> str:
+    os.makedirs(skill_dir, exist_ok=True)
+    skill_md_path = os.path.join(skill_dir, "SKILL.md")
+    final_description = description or "No description provided"
+    body = content or ""
+    markdown = (
+        f"---\n"
+        f"name: {skill_name}\n"
+        f"description: {final_description}\n"
+        f"---\n\n"
+        f"{body}\n"
+    )
+    with open(skill_md_path, "w", encoding="utf-8") as f:
+        f.write(markdown)
+    return skill_md_path
+
 def load_skills(project_id: Optional[int] = None) -> List[Dict[str, Any]]:
     data = _load_data()
     if project_id is not None:
@@ -205,9 +226,7 @@ async def upload_skill(
                 skill_name = os.path.splitext(filename)[0]
         
         # Create a safe directory name for the skill
-        import re
-        safe_name = re.sub(r'[^a-zA-Z0-9_\-]', '_', skill_name).lower()
-        if not safe_name: safe_name = "skill"
+        safe_name = _safe_skill_dir_name(skill_name)
         final_skill_id = f"{safe_name}_{datetime.now().strftime('%Y%m%d%H%M%S')}"
         final_skill_dir = os.path.join(SKILL_HUB_DIR, final_skill_id)
         
@@ -264,6 +283,15 @@ def create_skill(skill: SkillCreate):
     new_skill_dict = skill.dict()
     if not new_skill_dict.get("installation_time"):
         new_skill_dict["installation_time"] = datetime.now().strftime("%Y年%m月%d日")
+    if not new_skill_dict.get("file_path"):
+        skill_dir = os.path.join(SKILL_HUB_DIR, _safe_skill_dir_name(new_skill_dict["id"]))
+        _write_skill_markdown(
+            skill_dir=skill_dir,
+            skill_name=new_skill_dict["name"],
+            description=new_skill_dict.get("description"),
+            content=new_skill_dict.get("content", ""),
+        )
+        new_skill_dict["file_path"] = skill_dir
     
     data.append(new_skill_dict)
     _save_data(data)
@@ -279,6 +307,13 @@ def update_skill(skill_id: str, skill: SkillUpdate, project_id: Optional[int] =
             updated_item = item.copy()
             update_data = skill.dict(exclude_unset=True)
             updated_item.update(update_data)
+            if updated_item.get("file_path"):
+                _write_skill_markdown(
+                    skill_dir=updated_item["file_path"],
+                    skill_name=updated_item.get("name") or item.get("name") or "skill",
+                    description=updated_item.get("description"),
+                    content=updated_item.get("content", ""),
+                )
             data[i] = updated_item
             _save_data(data)
             return Skill(**updated_item)

backend/app/core/nanobot.py

@@ -1,6 +1,7 @@
 import asyncio
 import sys
 import os
+import shutil
 from pathlib import Path
 from typing import List, Callable, Awaitable, Any, Dict
 
@@ -47,6 +48,7 @@ class NanobotIntegration:
         # Set workspace path to backend/data/workspace
         workspace_path = Path(os.path.join(os.path.dirname(os.path.dirname(os.path.dirname(__file__))), "data", "workspace"))
         workspace_path.mkdir(parents=True, exist_ok=True)
+        self._sync_builtin_skills_to_workspace(workspace_path)
         
         # Override config workspace path via environment variable (since config is loaded from env)
         os.environ["NANOBOT_AGENTS__DEFAULTS__WORKSPACE"] = str(workspace_path)
@@ -87,6 +89,20 @@ class NanobotIntegration:
 
         self._register_custom_tools(self.agent)
 
+    def _sync_builtin_skills_to_workspace(self, workspace_path: Path) -> None:
+        builtin_root = Path(__file__).resolve().parents[1] / "skills_builtin"
+        workspace_skills_root = workspace_path / "skills"
+        workspace_skills_root.mkdir(parents=True, exist_ok=True)
+
+        for skill_name in ("nl2sql", "visualization"):
+            source_dir = builtin_root / skill_name
+            source_skill_file = source_dir / "SKILL.md"
+            if not source_skill_file.exists():
+                continue
+            target_dir = workspace_skills_root / skill_name
+            target_dir.mkdir(parents=True, exist_ok=True)
+            shutil.copy2(source_skill_file, target_dir / "SKILL.md")
+
     def _register_custom_tools(self, agent: AgentLoop):
         from app.tools.nl2sql import NL2SQLTool
         from app.tools.visualization import VisualizationTool

backend/app/schemas/chart.py

@@ -123,13 +123,6 @@ class ChartGenerationResponse(BaseModel):
     chart_type: Literal[
         "line", "multi_line", "bar", "pie", "grouped_bar", "stacked_bar", "area", ""
     ] = Field(..., description="The type of chart generated, or empty string if none")
-    chart_spec: Optional[Union[
-        LineChartSchema,
-        MultiLineChartSchema,
-        BarChartSchema,
-        PieChartSchema,
-        GroupedBarChartSchema,
-        StackedBarChartSchema,
-        AreaChartSchema
-    ]] = Field(None, description="The generated Vega-Lite chart specification")
+    # Using Dict[str, Any] allows LLM to output valid Vega-Lite spec directly, avoiding Pydantic strict model serialization issues with dynamic fields
+    chart_spec: Optional[Dict[str, Any]] = Field(None, description="The generated Vega-Lite chart specification")
     can_visualize: bool = Field(..., description="Whether the data can be visualized")

backend/app/skills_builtin/nl2sql/SKILL.md

@@ -0,0 +1,24 @@
+---
+description: Data Analysis and SQL Generation
+metadata:
+  nanobot:
+    always: true
+---
+
+# NL2SQL Data Analysis Skill
+
+You are an expert data analyst. You have access to a powerful `nl2sql` tool that can query the connected database using natural language.
+
+## When to use this skill
+- When the user asks to query, analyze, aggregate, or fetch data from the database.
+- Examples: "Show me the top 10 sales", "What is the average revenue by month?", "How many users registered yesterday?".
+
+## How to use this skill
+- Call the `nl2sql` tool with the user's natural language query.
+- If the user explicitly asks to "visualize" or "plot" the data in the SAME message as the query (e.g., "Show me sales by region and plot it as a pie chart"), you can set `generate_chart=True` in the `nl2sql` tool.
+- If the user ONLY asks to query data, set `generate_chart=False` (default).
+
+## After using the tool
+- The tool will return a summary of the executed query and a sample of the results.
+- Use this information to provide a clear, concise, and helpful response to the user.
+- If a chart was successfully generated by the tool, inform the user that the chart is available in the visualization panel.

backend/app/skills_builtin/visualization/SKILL.md

@@ -0,0 +1,23 @@
+---
+description: Data Visualization and Chart Generation
+metadata:
+  nanobot:
+    always: true
+---
+
+# Data Visualization Skill
+
+You are an expert data visualization specialist. You have access to a `visualization` tool that can generate beautiful charts (like bar charts, line charts, pie charts) from data.
+
+## When to use this skill
+- When the user asks to visualize, plot, or draw a chart based on data that has ALREADY been queried or is currently in context.
+- Examples: "Visualize it as a bar chart", "Plot the trend over time", "Draw a pie chart of the regions".
+- DO NOT use this tool if the data hasn't been queried yet. If the user asks a new question and wants it visualized (e.g., "Show me sales and plot it"), use the `nl2sql` tool with `generate_chart=True` instead, or call `nl2sql` first and then this tool.
+
+## How to use this skill
+- Call the `visualization` tool with the user's specific visualization request (e.g., "plot as a pie chart").
+- The tool relies on the data from the most recent SQL query. It will automatically read this data from the context.
+
+## After using the tool
+- The tool will return a success message and the reasoning for the chosen chart type.
+- Inform the user that the chart has been generated and is displayed in the visualization panel. Explain briefly what the chart shows if helpful.

backend/app/tools/nl2sql.py

@@ -1,6 +1,6 @@
 import json
 import logging
-from typing import Any, Dict
+from typing import Any
 
 from nanobot.agent.tools.base import Tool
 from app.agent.nl2sql import process_nl2sql, NL2SQLRequest, NL2SQLResponse
@@ -14,7 +14,7 @@ def _build_sql_chart_viz(nl2sql_result: NL2SQLResponse) -> dict:
     payload = {
         "sql": nl2sql_result.sql,
         "result": nl2sql_result.result,
-        "chart": chart.model_dump() if chart else None,
+        "chart": chart.model_dump(by_alias=True, exclude_none=True) if chart else None,
         "error": nl2sql_result.error,
     }
     return jsonable_encoder(payload)

backend/app/tools/visualization.py

@@ -60,7 +60,7 @@ class VisualizationTool(Tool):
                 viz_payload = {
                     "sql": existing_viz.get("sql", ""),
                     "result": data,
-                    "chart": chart_response.model_dump(),
+                    "chart": chart_response.model_dump(by_alias=True, exclude_none=True),
                     "error": None,
                 }
                 encoded_viz = jsonable_encoder(viz_payload)

frontend/src/components/ChatInterface.tsx

@@ -313,8 +313,8 @@ export function ChatInterface() {
   }): MessageViz => {
     const rows = Array.isArray(payload.result) ? payload.result : [];
     const chart = payload.chart ?? undefined;
-    const canVisualize = Boolean(chart?.can_visualize);
-    const chartSpec = canVisualize ? (chart?.chart_spec ?? null) : null;
+    const canVisualize = chart?.can_visualize ?? Boolean(chart?.chart_spec);
+    const chartSpec = chart?.chart_spec ?? null;
     return {
       sql: typeof payload.sql === "string" ? payload.sql : "",
       rows,
@@ -553,12 +553,12 @@ export function ChatInterface() {
       let renderedText = "";
 
       const flushAssistant = (force = false) => {
-        if (streamedText === renderedText) return;
+        if (streamedText === renderedText && !force) return;
         if (force) {
           renderedText = streamedText;
           setMessagesForSession(targetSessionKey, (prev) =>
             prev.map((msg) =>
-              msg.id === assistantId ? { ...msg, content: streamedText, awaitingFirstToken: false } : msg
+              msg.id === assistantId ? { ...msg, content: streamedText, awaitingFirstToken: false, viz: streamedViz ?? msg.viz } : msg
             )
           );
           return;
@@ -571,7 +571,7 @@ export function ChatInterface() {
           renderedText = streamedText;
           setMessagesForSession(targetSessionKey, (prev) =>
             prev.map((msg) =>
-              msg.id === assistantId ? { ...msg, content: streamedText, awaitingFirstToken: false } : msg
+              msg.id === assistantId ? { ...msg, content: streamedText, awaitingFirstToken: false, viz: streamedViz ?? msg.viz } : msg
             )
           );
         });
@@ -645,11 +645,7 @@ export function ChatInterface() {
           if (payload.type === "viz") {
             pushProgressLog("可视化结果已生成");
             streamedViz = buildMessageViz(payload);
-            setMessagesForSession(targetSessionKey, (prev) =>
-              prev.map((msg) =>
-                 msg.id === assistantId ? { ...msg, viz: streamedViz || undefined } : msg
-              )
-            );
+            flushAssistant(true); // 立即把 viz 状态刷入 messages
           }
          }
        }

frontend/src/components/InlineVisualizationCard.tsx

@@ -171,8 +171,8 @@ export function InlineVisualizationCard({ viz }: InlineVisualizationCardProps) {
         </div>
 
         {view === "chart" ? (
-          viz.canVisualize && viz.chartSpec && objectRows.length > 0 ? (
-            <div className="w-full h-80 rounded-xl border border-zinc-100 p-2">
+          viz.chartSpec && objectRows.length > 0 ? (
+            <div className="w-full h-80 min-h-[320px] rounded-xl border border-zinc-100 p-2">
               <VegaChart data={objectRows} spec={viz.chartSpec} />
             </div>
           ) : (

frontend/src/components/VegaChart.tsx

@@ -30,23 +30,27 @@ export const VegaChart: React.FC<VegaChartProps> = ({ data, spec }) => {
   }, []);
 
   const vegaSpec: any = useMemo(() => {
-    // Clone spec and ensure tooltip is enabled in mark if not already specified
-    const baseSpec = { ...spec };
+    // Deep clone spec to avoid mutating React state/props
+    const baseSpec = JSON.parse(JSON.stringify(spec));
+    
+    // Ensure tooltip is enabled in mark if not already specified
     if (typeof baseSpec.mark === 'string') {
       baseSpec.mark = { type: baseSpec.mark, tooltip: true };
     } else if (typeof baseSpec.mark === 'object' && baseSpec.mark !== null) {
-      baseSpec.mark = { ...baseSpec.mark, tooltip: true };
+      baseSpec.mark.tooltip = true;
     }
 
     // Add highlight effect: hover over an element makes others transparent
     // 1. Define hover param
     if (!baseSpec.params) {
-      baseSpec.params = [
-        {
+      baseSpec.params = [];
+    }
+    const hasHighlight = baseSpec.params.some((p: any) => p.name === "highlight");
+    if (!hasHighlight) {
+      baseSpec.params.push({
         name: "highlight",
         select: { type: "point", on: "mouseover", clear: "mouseout" }
-        }
-      ];
+      });
     }
 
     // 2. Add conditional opacity to encoding
@@ -64,7 +68,7 @@ export const VegaChart: React.FC<VegaChartProps> = ({ data, spec }) => {
 
     // Also add cursor: pointer for marks
     if (typeof baseSpec.mark === 'object' && baseSpec.mark !== null) {
-      (baseSpec.mark as any).cursor = "pointer";
+      baseSpec.mark.cursor = "pointer";
     }
 
     return {
@@ -77,12 +81,17 @@ export const VegaChart: React.FC<VegaChartProps> = ({ data, spec }) => {
     };
   }, [data, size.height, size.width, spec]);
 
+  const handleError = (error: any) => {
+    console.error("VegaEmbed rendering error:", error, "Spec:", vegaSpec);
+  };
+
   return (
-    <div className="w-full h-full" ref={containerRef}>
+    <div className="w-full h-full min-h-[300px]" ref={containerRef}>
       <VegaEmbed 
         spec={vegaSpec} 
         options={{ actions: false }} 
         style={{width: '100%', height: '100%'}} 
+        onError={handleError}
       />
     </div>
   );

nanobot/nanobot/providers/litellm_provider.py

@@ -214,6 +214,8 @@ class LiteLLMProvider(LLMProvider):
         max_tokens: int = 4096,
         temperature: float = 0.7,
         reasoning_effort: str | None = None,
+        request_timeout: float | None = None,
+        num_retries: int | None = None,
     ) -> LLMResponse:
         """
         Send a chat completion request via LiteLLM.
@@ -269,6 +271,12 @@ class LiteLLMProvider(LLMProvider):
             kwargs["tools"] = tools
             kwargs["tool_choice"] = "auto"
         
+        if request_timeout is not None:
+            kwargs["timeout"] = request_timeout
+
+        if num_retries is not None:
+            kwargs["num_retries"] = max(0, int(num_retries))
+
         try:
             response = await acompletion(**kwargs)
             return self._parse_response(response)