added clear cache tool

README.md

@@ -22,7 +22,7 @@ Gradio-based MCP server for Ragmint, enabling **Retrieval-Augmented Generation (
 
 ## 🧩 Overview
 
-Ragmint MCP Server exposes the full power of **Ragmint**, a modular Python library for **evaluating, optimizing, and tuning RAG pipelines**, through a **Multimodal Control Plane (MCP)**. This allows external clients (like Claude Desktop or Cursor) to **run experiments, retrieve leaderboard results, and tune RAG parameters programmatically**.
+Ragmint MCP Server exposes the full power of **Ragmint**, a modular Python library for **evaluating, optimizing, and tuning RAG pipelines**, through a **Multimodal Control Plane (MCP)**. This allows external clients (like Claude Desktop or Cursor) to **run experiments and tune RAG parameters programmatically**.
 
 ## Ragmint
 
@@ -78,17 +78,8 @@ export GOOGLE_API_KEY="your_gemini_key"
 
 ## 🧠 MCP Usage
 
-Ragmint MCP Server provides Python-callable interfaces for programmatic control. Example usage with MCP:
+Ragmint MCP Server provides Python-callable interfaces for programmatic control. You can find an example of MCP usage in the [Ragmint MCP Server Space](https://huggingface.co/spaces/andyolivers/ragmint-mcp-server) on Hugging Face.
 
-```python
-from mcp_client import MCPClient
-
-client = MCPClient(server_url="http://localhost:7860")
-
-# Run Auto-RAG tuning
-config, results = client.autotune(docs_path="data/docs/", trials=5)
-print("Best config:", config)
-```
 
 ---
 
@@ -142,6 +133,22 @@ ragmint_mcp_server/
 ├── models.py
 └── api.py
 ```
+---
+## 🔧 MCP Tools (app.py)
+
+The `app.py` file provides the Gradio UI and also registers the functions exposed as **MCP Tools**, enabling external MCP clients (Claude Desktop, Cursor, VS Code MCP extension, etc.) to call Ragmint programmatically.
+
+`app.py` launches the FastAPI backend (`api.py`) in a background thread and exposes the following MCP tools:
+
+| MCP Tool  | Python Function        | Description                                                                        |
+|-----------|------------------------|------------------------------------------------------------------------------------|
+| upload_docs | upload_docs_tool()     | Uploads `.txt` files or remote URLs into the configured `docs_path`.              |
+| upload_urls | upload_urls_tool()     | Downloads remote files from external URLs and stores them inside `docs_path`.     |
+| optimize_rag | optimize_rag_tool()    | Runs explicit hyperparameter optimization for a RAG pipeline.                     |
+| autotune  | autotune_tool()        | Automatically recommends best chunking + embedding configuration.                 |
+| generate_qa | generate_qa_tool()     | Generates synthetic QA validation dataset for evaluation.                         |
+| clear_cache | clear_cache_tool()     | Deletes all docs inside `data/docs` to reset the workspace.                       |
+
 ---
 
 ## 📥 Inputs
@@ -149,21 +156,35 @@ ragmint_mcp_server/
 The Ragmint MCP Server exposes three main endpoints with the following inputs:
 
 
-### 1. Upload Documents
+### 1. Upload Documents (`upload_docs`)
 
-Input: `.txt` files to upload to the documents directory (`docs_path`).
+Input: `.txt` files or file-like objects to upload to the documents directory (`docs_path`).
 
 <details>
 <summary>View Input Model</summary>
 
 | Field | Type | Description | Example |
-|-------|------|-------------|---------|
-| file | File | Text file to be processed | sample.txt |
+|--------|-------|-------------|---------|
+| files | File[] | Local `.txt` files selected or passed from MCP client | ["sample.txt"] |
 | docs_path | str | Directory where files are stored | data/docs |
+</details>
+
+
+### 2. Upload URLs (`upload_urls`)
+
+Input: List of URLs referencing `.txt` files to download and store in `docs_path`.
+
+<details>
+<summary>View Input Model</summary>
+
+| Field | Type | Description | Example |
+|--------|-------|-------------|---------|
+| urls | List[str] | List of URLs pointing to remote documents | ["https://example.com/doc.txt"] |
+| docs_path | str | Directory where downloaded files are saved | data/docs |
 
 </details>
 
-### 2. Optimize RAG
+### 3. Optimize RAG (`optimize_rag`)
 
 Input: JSON object following the `OptimizeRequest` model.
 
@@ -187,7 +208,7 @@ Input: JSON object following the `OptimizeRequest` model.
 
 </details>
 
-### 3. Autotune RAG
+### 4. Autotune RAG (`autotune`)
 
 Input: JSON object following the `AutotuneRequest` model.
 
@@ -207,7 +228,7 @@ Input: JSON object following the `AutotuneRequest` model.
 
 </details>
 
-### 4. Generate QA
+### 5. Generate QA (`generate_qa`)
 
 Input: JSON object following the `QARequest` model.
 <details>
@@ -223,13 +244,26 @@ Input: JSON object following the `QARequest` model.
 
 </details>
 
+### 6. Clear Cache (`clear_cache`)
+
+Deletes all stored documents from `data/docs`.
+
+<details>
+<summary>View Input Model</summary>
+
+| Field | Type | Description | Example |
+|--------|-------|-------------|---------|
+| docs_path | str | Folder to wipe clean | data/docs |
+
+</details>
+
 ---
 
 ## 📤 Outputs
 
 The Ragmint MCP Server exposes three main endpoints with the following example outputs:
 
-### 1. Upload Documents Response
+### 1. Upload Documents Response (`upload_docs`)
 
 <details>
 <summary>View Response Example</summary>
@@ -251,7 +285,28 @@ The Ragmint MCP Server exposes three main endpoints with the following example o
 ✅ Confirms your documents are ready for RAG operations.
 
 
-### 2. Optimize RAG Response
+### 2. Upload URLs Response (`upload_urls`)
+
+<details>
+<summary>View Response Example</summary>
+
+```json
+{
+  "status": "ok",
+  "uploaded_files": ["doc.txt"],
+  "docs_path": "data/docs"
+}
+```
+</details> 
+
+- **status**: `"ok"` → Indicates that the upload was successful.
+- **uploaded_files**: List of file names that were successfully uploaded.
+- **docs_path**: The directory where the uploaded documents are stored.
+
+✅ Confirms your documents are ready for RAG operations.
+
+
+### 3. Optimize RAG Response (`optimize_rag`)
 
 <details>
 <summary>View Response Example</summary>
@@ -312,7 +367,7 @@ The Ragmint MCP Server exposes three main endpoints with the following example o
   - **corpus_size** → Total size in characters or tokens.
 
 
-### 3. Autotune RAG Response
+### 4. Autotune RAG Response (`autotune`)
 
 <details>
 <summary>View Response Example</summary>
@@ -383,7 +438,7 @@ The Ragmint MCP Server exposes three main endpoints with the following example o
 🧠 **Difference from Optimize**: Autotune automatically selects the best hyperparameters, rather than testing all user-specified combinations.
 
 
-### 4. Generate QA Response
+### 5. Generate QA Response (`generate_qa`)
 
 <details>
 <summary>View Response Example</summary>
@@ -419,6 +474,24 @@ The Ragmint MCP Server exposes three main endpoints with the following example o
   - **expected_answer** → The reference answer corresponding to that question.
 - **status**: `"finished"` → QA generation completed successfully.
 
+
+### 6. Clear Cache Response (`clear_cache`)
+
+<details>
+<summary>View Response Example</summary>
+
+```json
+{
+  "status": "ok",
+  "deleted_files": 7,
+  "docs_path": "data/docs"
+}
+```
+</details>
+
+- **deleted_files**: Number of documents removed
+- **status**: "ok" indicates successful workspace reset
+
 ---
 
 ## 📘 License

api.py

@@ -282,6 +282,42 @@ def generate_validation_qa_endpoint(req: QARequest):
         raise HTTPException(status_code=500, detail=str(exc))
 
 
+@app.post("/clear_cache")
+async def clear_cache(docs_path: str = Form(DEFAULT_DATA_DIR)):
+    """
+    Delete all files inside docs_path but keep the directory.
+    Useful to reset uploaded documents for RAG runs.
+    """
+    if not os.path.exists(docs_path):
+        raise HTTPException(status_code=400, detail=f"docs_path does not exist: {docs_path}")
+
+    removed = []
+    for root, dirs, files in os.walk(docs_path, topdown=False):
+        for name in files:
+            file_path = os.path.join(root, name)
+            try:
+                os.remove(file_path)
+                removed.append(name)
+            except Exception as e:
+                logger.error(f"Failed to remove {file_path}: {e}")
+
+        for name in dirs:
+            dir_path = os.path.join(root, name)
+            try:
+                shutil.rmtree(dir_path)
+                removed.append(f"{name}/")
+            except Exception as e:
+                logger.error(f"Failed to remove {dir_path}: {e}")
+
+    return {
+        "status": "cleared",
+        "docs_path": docs_path,
+        "removed_items": removed,
+        "total_removed": len(removed),
+    }
+
+
+
 def start_api():
     import uvicorn
     uvicorn.run(app, host="0.0.0.0", port=8000, log_level="info")

app.py

@@ -23,6 +23,25 @@ def call_api(endpoint: str, payload: dict) -> str:
         return str(e)
 
 
+def clear_cache_tool(docs_path="data/docs"):
+    """
+    🧹 Clear Cache MCP Tool
+    Deletes all files and directories inside docs_path on the server.
+    Accepts:
+    - local paths (str), default='data/docs/'
+    """
+    try:
+        r = requests.post(
+            f"{BASE_INTERNAL}/clear_cache",
+            data={"docs_path": docs_path},
+            timeout=60
+        )
+        r.raise_for_status()
+        return r.json()
+    except Exception as e:
+        return {"error": str(e)}
+
+
 def upload_docs_tool(files, docs_path="data/docs"):
     """
     Upload documents to the server's docs folder via FastAPI /upload_docs.
@@ -136,11 +155,11 @@ with gr.Blocks(theme=gr.themes.Soft()) as demo:
 
     # Upload MCP Documents (no file uploader)
     with gr.Column():
-        gr.Markdown("## Upload Documents via MCP")
-        gr.Markdown("📂 Upload files (local paths or URLs) to your `data/docs` folder on MCP.")
+        gr.Markdown("## Upload Documents (URLs) via MCP")
+        gr.Markdown("📂 Upload files (URLs) to your `data/docs` folder on MCP.")
         upload_mcp_input = gr.Textbox(
             lines=5,
-            placeholder='Enter list of files/URLs, e.g., ["file1.txt","file2.pdf"]',
+            placeholder='Enter list of URLs (e.g., ["https://example.com/example.txt",...])',
             label="Files (JSON list)"
         )
         upload_mcp_path = gr.Textbox(value=DEFAULT_UPLOAD_PATH, label="Docs Path")
@@ -198,6 +217,16 @@ with gr.Blocks(theme=gr.themes.Soft()) as demo:
         qa_btn.click(generate_qa_tool, inputs=qa_input, outputs=qa_out)
         gr.Markdown("---")
 
+    # Clear Cache
+    with gr.Column():
+        gr.Markdown("## Clear Cache")
+        gr.Markdown("🧹 Deletes all files and directories inside docs_path on the server.")
+        clear_path = gr.Textbox(value=DEFAULT_UPLOAD_PATH, label="Docs Path to Clear")
+        clear_btn = gr.Button("Clear Cache", variant="primary")
+        clear_out = gr.JSON(label="Response")
+        clear_btn.click(clear_cache_tool, inputs=[clear_path], outputs=clear_out)
+        gr.Markdown("---")
+
 if __name__ == "__main__":
 
     demo.launch(

data/docs/cisco.txt

@@ -1,7 +0,0 @@
-Cisco Systems, Inc. is a global technology company headquartered in San Jose, California, specializing in networking hardware, software, and telecommunications equipment.
-Founded in 1984 by Leonard Bosack and Sandy Lerner, Cisco initially focused on connecting computers at Stanford University, pioneering modern network routing technology.
-Cisco is widely recognized for its enterprise routers, switches, and networking solutions, which form the backbone of many corporate and service provider networks worldwide.
-The company also offers cybersecurity solutions, including firewalls, VPNs, and intrusion prevention systems, to protect enterprises from evolving digital threats.
-Cisco has a strong presence in cloud computing, collaboration tools, and Internet of Things (IoT) technology, helping organizations modernize and automate their IT infrastructure.
-The company invests heavily in research and development, maintaining innovation in areas such as artificial intelligence (AI), software-defined networking (SDN), and network automation.
-Cisco’s stock is publicly traded on the NASDAQ under the ticker symbol CSCO, and it consistently ranks among the largest technology companies globally in revenue and market capitalization.