cite-agent 1.3.4__py3-none-any.whl → 1.3.6__py3-none-any.whl

cite_agent/__version__.py

@@ -1 +1 @@
-__version__ = "1.3.4"
+__version__ = "1.3.6"

cite_agent/cli.py

@@ -256,6 +256,19 @@ class NocturnalCLI:
         if not await self.initialize():
             return
         
+        # Detect if user is in a project directory (R, Python, Node, Jupyter, etc.)
+        try:
+            from .project_detector import ProjectDetector
+            detector = ProjectDetector()
+            project_info = detector.detect_project()
+            
+            if project_info:
+                # Show project banner
+                banner = detector.format_project_banner(project_info)
+                self.console.print(banner, style="dim")
+        except:
+            pass  # Silently skip if detection fails
+        
         self.console.print("\n[bold]🤖 Interactive Mode[/] — Type your questions or 'quit' to exit")
         self.console.rule(style="magenta")
         

cite_agent/enhanced_ai_agent.py

@@ -1698,10 +1698,55 @@ class EnhancedNocturnalAgent:
                 
                 elif response.status == 503:
                     # Backend AI service temporarily unavailable (Cerebras/Groq rate limited)
-                    # Graceful degradation: return helpful message instead of error
+                    # Auto-retry silently with exponential backoff
+                    
+                    print("\n💭 Thinking... (backend is busy, retrying automatically)")
+                    
+                    import asyncio
+                    retry_delays = [5, 15, 30]  # Exponential backoff
+                    
+                    for retry_num, delay in enumerate(retry_delays):
+                        await asyncio.sleep(delay)
+                        
+                        # Retry the request
+                        async with self.session.post(url, json=payload, headers=headers, timeout=60) as retry_response:
+                            if retry_response.status == 200:
+                                # Success!
+                                data = await retry_response.json()
+                                response_text = data.get('response', '')
+                                tokens = data.get('tokens_used', 0)
+                                
+                                all_tools = tools_used or []
+                                all_tools.append("backend_llm")
+                                
+                                self.workflow.save_query_result(
+                                    query=query,
+                                    response=response_text,
+                                    metadata={
+                                        "tools_used": all_tools,
+                                        "tokens_used": tokens,
+                                        "model": data.get('model'),
+                                        "provider": data.get('provider'),
+                                        "retries": retry_num + 1
+                                    }
+                                )
+                                
+                                return ChatResponse(
+                                    response=response_text,
+                                    tokens_used=tokens,
+                                    tools_used=all_tools,
+                                    model=data.get('model', 'llama-3.3-70b'),
+                                    timestamp=data.get('timestamp', datetime.now(timezone.utc).isoformat()),
+                                    api_results=api_results
+                                )
+                            elif retry_response.status != 503:
+                                # Different error, stop retrying
+                                break
+                    
+                    # All retries exhausted
                     return ChatResponse(
-                        response="⚠️ AI service temporarily busy. Try again in a moment, or rephrase your question.",
-                        error_message="Service temporarily unavailable"
+                        response="❌ Service unavailable. Please try again in a few minutes.",
+                        error_message="Service unavailable after retries"
                     )
                 
                 elif response.status == 200:
@@ -2619,7 +2664,8 @@ class EnhancedNocturnalAgent:
             # Quick check if query might need shell
             question_lower = request.question.lower()
             might_need_shell = any(word in question_lower for word in [
-                'directory', 'folder', 'where', 'find', 'list', 'files', 'look', 'search', 'check', 'into'
+                'directory', 'folder', 'where', 'find', 'list', 'files', 'look', 'search', 'check', 'into',
+                'show', 'open', 'read', 'display', 'cat', 'view', 'contents', '.r', '.py', '.csv', '.ipynb'
             ])
             
             if might_need_shell and self.shell_session:
@@ -2631,19 +2677,28 @@ Previous conversation: {json.dumps(self.conversation_history[-2:]) if self.conve
 
 Respond ONLY with JSON:
 {{
-  "action": "pwd|ls|find|none",
+  "action": "pwd|ls|find|read_file|none",
   "search_target": "cm522" (if find),
   "search_path": "~/Downloads" (if find),
-  "target_path": "/full/path" (if ls on previous result)
+  "target_path": "/full/path" (if ls on previous result),
+  "file_path": "/full/path/to/file.R" (if read_file)
 }}
 
 Examples:
 "where am i?" → {{"action": "pwd"}}
 "what files here?" → {{"action": "ls"}}
-"find cm522 in downloads" → {{"action": "find", "search_target": "cm522", "search_path": "~/Downloads"}}
-"look into it" + Previous: "Found /path/to/dir" → {{"action": "ls", "target_path": "/path/to/dir"}}
+"find cm522" → {{"action": "find", "search_target": "cm522"}}
+"look into it" + Previous: "Found /path" → {{"action": "ls", "target_path": "/path"}}
+"show me calculate_betas.R" → {{"action": "read_file", "file_path": "calculate_betas.R"}}
+"open regression.R" → {{"action": "read_file", "file_path": "regression.R"}}
+"read that file" + Previous: "regression.R" → {{"action": "read_file", "file_path": "regression.R"}}
+"display analysis.py" → {{"action": "read_file", "file_path": "analysis.py"}}
+"cat data.csv" → {{"action": "read_file", "file_path": "data.csv"}}
+"what columns does it have?" + Previous: file was shown → {{"action": "none"}} (LLM will parse from conversation)
 "Tesla revenue" → {{"action": "none"}}
 
+KEY: If query mentions a specific FILENAME (*.R, *.py, *.csv), use read_file, NOT find!
+
 JSON:"""
 
                 try:
@@ -2701,6 +2756,65 @@ JSON:"""
                                     "search_results": f"No directories matching '{search_target}' found in {search_path}"
                                 }
                             tools_used.append("shell_execution")
+                    
+                    elif shell_action == "read_file":
+                        # NEW: Read and inspect file (R, Python, CSV, etc.)
+                        import re  # Import at function level
+                        
+                        file_path = plan.get("file_path", "")
+                        if not file_path and might_need_shell:
+                            # Try to infer from query (e.g., "show me calculate_betas.R")
+                            filenames = re.findall(r'([a-zA-Z0-9_-]+\.[a-zA-Z]{1,4})', request.question)
+                            if filenames:
+                                # Check if file exists in current directory
+                                pwd = self.execute_command("pwd").strip()
+                                file_path = f"{pwd}/{filenames[0]}"
+                        
+                        if file_path:
+                            if debug_mode:
+                                print(f"🔍 READING FILE: {file_path}")
+                            
+                            # Read file content (first 100 lines to detect structure)
+                            cat_output = self.execute_command(f"head -100 {file_path}")
+                            
+                            if not cat_output.startswith("ERROR"):
+                                # Detect file type and extract structure
+                                file_ext = file_path.split('.')[-1].lower()
+                                
+                                # Extract column/variable info based on file type
+                                columns_info = ""
+                                if file_ext in ['csv', 'tsv']:
+                                    # CSV: first line is usually headers
+                                    first_line = cat_output.split('\n')[0] if cat_output else ""
+                                    columns_info = f"CSV columns: {first_line}"
+                                elif file_ext in ['r', 'rmd']:
+                                    # R script: look for dataframe column references (df$columnname)
+                                    column_refs = re.findall(r'\$(\w+)', cat_output)
+                                    unique_cols = list(dict.fromkeys(column_refs))[:10]
+                                    if unique_cols:
+                                        columns_info = f"Detected columns/variables: {', '.join(unique_cols)}"
+                                elif file_ext == 'py':
+                                    # Python: look for DataFrame['column'] or df.column
+                                    column_refs = re.findall(r'\[[\'""](\w+)[\'"]\]|\.(\w+)', cat_output)
+                                    unique_cols = list(dict.fromkeys([c[0] or c[1] for c in column_refs if c[0] or c[1]]))[:10]
+                                    if unique_cols:
+                                        columns_info = f"Detected columns/attributes: {', '.join(unique_cols)}"
+                                
+                                api_results["file_context"] = {
+                                    "file_path": file_path,
+                                    "file_type": file_ext,
+                                    "content_preview": cat_output[:2000],  # First 2000 chars
+                                    "structure": columns_info,
+                                    "full_content": cat_output  # Full content for analysis
+                                }
+                                tools_used.append("file_read")
+                                
+                                if debug_mode:
+                                    print(f"🔍 FILE STRUCTURE: {columns_info}")
+                            else:
+                                api_results["file_context"] = {
+                                    "error": f"Could not read file: {file_path}"
+                                }
                 
                 except Exception as e:
                     if debug_mode:

cite_agent/project_detector.py

@@ -0,0 +1,148 @@
+#!/usr/bin/env python3
+"""
+Generic project detection - works with ANY IDE/project type
+Not RStudio-specific - detects R, Python, Node, Jupyter, etc.
+"""
+
+import os
+from pathlib import Path
+from typing import Dict, List, Optional, Any
+import glob
+
+
+class ProjectDetector:
+    """Detects project type and provides context"""
+    
+    def __init__(self, working_dir: Optional[str] = None):
+        self.working_dir = Path(working_dir or os.getcwd())
+    
+    def detect_project(self) -> Optional[Dict[str, Any]]:
+        """
+        Detect what kind of project user is working in
+        Returns project info or None if not in a project
+        """
+        project_info = {
+            "type": None,
+            "name": None,
+            "recent_files": [],
+            "description": None
+        }
+        
+        # Check for R project (.Rproj file OR 2+ .R files)
+        rproj_files = list(self.working_dir.glob("*.Rproj"))
+        r_files = list(self.working_dir.glob("*.R")) + list(self.working_dir.glob("*.Rmd"))
+        
+        if rproj_files:
+            project_info["type"] = "R"
+            project_info["name"] = rproj_files[0].stem
+            project_info["recent_files"] = self._get_recent_files([".R", ".Rmd", ".qmd"])
+            project_info["description"] = f"R/RStudio project: {project_info['name']}"
+            return project_info
+        elif len(r_files) >= 2:
+            # R project without .Rproj file
+            project_info["type"] = "R"
+            project_info["name"] = self.working_dir.name
+            project_info["recent_files"] = self._get_recent_files([".R", ".Rmd", ".qmd"])
+            project_info["description"] = f"R project: {project_info['name']}"
+            return project_info
+        
+        # Check for Python project
+        if (self.working_dir / "pyproject.toml").exists() or \
+           (self.working_dir / "setup.py").exists() or \
+           (self.working_dir / "requirements.txt").exists():
+            project_info["type"] = "Python"
+            project_info["name"] = self.working_dir.name
+            project_info["recent_files"] = self._get_recent_files([".py", ".ipynb"])
+            project_info["description"] = f"Python project: {project_info['name']}"
+            return project_info
+        
+        # Check for Node.js project
+        if (self.working_dir / "package.json").exists():
+            project_info["type"] = "Node"
+            project_info["name"] = self.working_dir.name
+            project_info["recent_files"] = self._get_recent_files([".js", ".ts", ".jsx", ".tsx"])
+            project_info["description"] = f"Node.js project: {project_info['name']}"
+            return project_info
+        
+        # Check for Jupyter/Data Science directory
+        ipynb_files = list(self.working_dir.glob("*.ipynb"))
+        if len(ipynb_files) >= 2:  # 2+ notebooks = likely data science project
+            project_info["type"] = "Jupyter"
+            project_info["name"] = self.working_dir.name
+            project_info["recent_files"] = self._get_recent_files([".ipynb", ".py", ".csv"])
+            project_info["description"] = f"Jupyter/Data Science project: {project_info['name']}"
+            return project_info
+        
+        # Check for Git repository
+        if (self.working_dir / ".git").exists():
+            project_info["type"] = "Git"
+            project_info["name"] = self.working_dir.name
+            # Get recent files of any code type
+            project_info["recent_files"] = self._get_recent_files([".py", ".js", ".R", ".java", ".cpp", ".rs"])
+            project_info["description"] = f"Git repository: {project_info['name']}"
+            return project_info
+        
+        # Not in a recognized project
+        return None
+    
+    def _get_recent_files(self, extensions: List[str], limit: int = 5) -> List[Dict[str, Any]]:
+        """Get recently modified files with given extensions"""
+        files = []
+        
+        for ext in extensions:
+            for filepath in self.working_dir.glob(f"**/*{ext}"):
+                if filepath.is_file() and not any(part.startswith('.') for part in filepath.parts):
+                    try:
+                        mtime = filepath.stat().st_mtime
+                        files.append({
+                            "name": filepath.name,
+                            "path": str(filepath),
+                            "relative_path": str(filepath.relative_to(self.working_dir)),
+                            "modified": mtime,
+                            "extension": ext
+                        })
+                    except:
+                        pass
+        
+        # Sort by modification time, newest first
+        files.sort(key=lambda f: f["modified"], reverse=True)
+        return files[:limit]
+    
+    def format_project_banner(self, project_info: Dict[str, Any]) -> str:
+        """Format project info as a nice banner"""
+        if not project_info or not project_info["type"]:
+            return ""
+        
+        icon_map = {
+            "R": "📊",
+            "Python": "🐍",
+            "Node": "📦",
+            "Jupyter": "📓",
+            "Git": "📂"
+        }
+        
+        icon = icon_map.get(project_info["type"], "📁")
+        
+        banner = f"\n{icon} {project_info['description']}\n"
+        
+        if project_info["recent_files"]:
+            banner += "📄 Recent files:\n"
+            for f in project_info["recent_files"][:3]:
+                banner += f"   • {f['relative_path']}\n"
+        
+        return banner
+    
+    def get_project_context_for_llm(self, project_info: Dict[str, Any]) -> str:
+        """Get project context to add to LLM prompts"""
+        if not project_info or not project_info["type"]:
+            return ""
+        
+        context = f"User is working in a {project_info['type']} project: {project_info['name']}\n"
+        
+        if project_info["recent_files"]:
+            context += "Recent files:\n"
+            for f in project_info["recent_files"][:5]:
+                context += f"- {f['relative_path']}\n"
+        
+        return context
+

{cite_agent-1.3.4.dist-info → cite_agent-1.3.6.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: cite-agent
-Version: 1.3.4
+Version: 1.3.6
 Summary: Terminal AI assistant for academic research with citation verification
 Home-page: https://github.com/Spectating101/cite-agent
 Author: Cite-Agent Team

{cite_agent-1.3.4.dist-info → cite_agent-1.3.6.dist-info}/RECORD

@@ -1,17 +1,18 @@
 cite_agent/__init__.py,sha256=wAXV2v8nNOmIAd0rh8196ItBl9hHWBVOBl5Re4VB77I,1645
 cite_agent/__main__.py,sha256=6x3lltwG-iZHeQbN12rwvdkPDfd2Rmdk71tOOaC89Mw,179
-cite_agent/__version__.py,sha256=U6E-HRsrit7kgSGgIeTI2eMUeyUCny5DH8LDV4I1o0g,22
+cite_agent/__version__.py,sha256=5ZbAQtod5QalTI1C2N07edlxplzG_Q2XvGOSyOok4uA,22
 cite_agent/account_client.py,sha256=yLuzhIJoIZuXHXGbaVMzDxRATQwcy-wiaLnUrDuwUhI,5725
 cite_agent/agent_backend_only.py,sha256=H4DH4hmKhT0T3rQLAb2xnnJVjxl3pOZaljL9r6JndFY,6314
 cite_agent/ascii_plotting.py,sha256=lk8BaECs6fmjtp4iH12G09-frlRehAN7HLhHt2crers,8570
 cite_agent/auth.py,sha256=YtoGXKwcLkZQbop37iYYL9BzRWBRPlt_D9p71VGViS4,9833
 cite_agent/backend_only_client.py,sha256=WqLF8x7aXTro2Q3ehqKMsdCg53s6fNk9Hy86bGxqmmw,2561
-cite_agent/cli.py,sha256=f0x19N4MB85XgfNtOPEtovLQO_qzTHsli5IpG8wra30,35029
+cite_agent/cli.py,sha256=Qq0Gt7sNVR4R2ue10KFZElOPYrujBG9xTOqy2qetxL4,35562
 cite_agent/cli_conversational.py,sha256=RAmgRNRyB8gQ8QLvWU-Tt23j2lmA34rQNT5F3_7SOq0,11141
 cite_agent/cli_enhanced.py,sha256=EAaSw9qtiYRWUXF6_05T19GCXlz9cCSz6n41ASnXIPc,7407
 cite_agent/cli_workflow.py,sha256=4oS_jW9D8ylovXbEFdsyLQONt4o0xxR4Xatfcc4tnBs,11641
 cite_agent/dashboard.py,sha256=VGV5XQU1PnqvTsxfKMcue3j2ri_nvm9Be6O5aVays_w,10502
-cite_agent/enhanced_ai_agent.py,sha256=HfCotlyrtBRXx6ENa8pRnG_x0JZ5izoIE87VwgLWAhU,165207
+cite_agent/enhanced_ai_agent.py,sha256=hOL17pDKQdD1MJZRXkjEBlqyNmTdA_pcxIkQEojysFM,172282
+cite_agent/project_detector.py,sha256=fPl5cLTy_oyufqrQ7RJ5IRVdofZoPqDRaQXW6tRtBJc,6086
 cite_agent/rate_limiter.py,sha256=-0fXx8Tl4zVB4O28n9ojU2weRo-FBF1cJo9Z5jC2LxQ,10908
 cite_agent/session_manager.py,sha256=B0MXSOsXdhO3DlvTG7S8x6pmGlYEDvIZ-o8TZM23niQ,9444
 cite_agent/setup_config.py,sha256=3m2e3gw0srEWA0OygdRo64r-8HK5ohyXfct0c__CF3s,16817
@@ -22,7 +23,7 @@ cite_agent/updater.py,sha256=udoAAN4gBKAvKDV7JTh2FJO_jIhNk9bby4x6n188MEY,8458
 cite_agent/web_search.py,sha256=FZCuNO7MAITiOIbpPbJyt2bzbXPzQla-9amJpnMpW_4,6520
 cite_agent/workflow.py,sha256=a0YC0Mzz4or1C5t2gZcuJBQ0uMOZrooaI8eLu2kkI0k,15086
 cite_agent/workflow_integration.py,sha256=A9ua0DN5pRtuU0cAwrUTGvqt2SXKhEHQbrHx16EGnDM,10910
-cite_agent-1.3.4.dist-info/licenses/LICENSE,sha256=XJkyO4IymhSUniN1ENY6lLrL2729gn_rbRlFK6_Hi9M,1074
+cite_agent-1.3.6.dist-info/licenses/LICENSE,sha256=XJkyO4IymhSUniN1ENY6lLrL2729gn_rbRlFK6_Hi9M,1074
 src/__init__.py,sha256=0eEpjRfjRjOTilP66y-AbGNslBsVYr_clE-bZUzsX7s,40
 src/services/__init__.py,sha256=pTGLCH_84mz4nGtYMwQES5w-LzoSulUtx_uuNM6r-LA,4257
 src/services/simple_enhanced_main.py,sha256=IJoOplCqcVUg3GvN_BRyAhpGrLm_WEPy2jmHcNCY6R0,9257
@@ -49,8 +50,8 @@ src/services/research_service/synthesizer.py,sha256=lCcu37PWhWVNphHKaJJDIC-JQ5OI
 src/services/search_service/__init__.py,sha256=UZFXdd7r6wietQ2kESXEyGffdfBbpghquecQde7auF4,137
 src/services/search_service/indexer.py,sha256=u3-uwdAfmahWWsdebDF9i8XIyp7YtUMIHzlmBLBnPPM,7252
 src/services/search_service/search_engine.py,sha256=S9HqQ_mk-8W4d4MUOgBbEGQGV29-eSuceSFvVb4Xk-k,12500
-cite_agent-1.3.4.dist-info/METADATA,sha256=AFuXC6Gvnkb1wUAuau1XbuD-xNVlJdmjoKBHyaG-0OM,12231
-cite_agent-1.3.4.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-cite_agent-1.3.4.dist-info/entry_points.txt,sha256=bJ0u28nFIxQKH1PWQ2ak4PV-FAjhoxTC7YADEdDenFw,83
-cite_agent-1.3.4.dist-info/top_level.txt,sha256=TgOFqJTIy8vDZuOoYA2QgagkqZtfhM5Acvt_IsWzAKo,15
-cite_agent-1.3.4.dist-info/RECORD,,
+cite_agent-1.3.6.dist-info/METADATA,sha256=Nw2biNkpNAmCACEe_2dJ5dGy-KvL7DtxfFuNhgNWIN4,12231
+cite_agent-1.3.6.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+cite_agent-1.3.6.dist-info/entry_points.txt,sha256=bJ0u28nFIxQKH1PWQ2ak4PV-FAjhoxTC7YADEdDenFw,83
+cite_agent-1.3.6.dist-info/top_level.txt,sha256=TgOFqJTIy8vDZuOoYA2QgagkqZtfhM5Acvt_IsWzAKo,15
+cite_agent-1.3.6.dist-info/RECORD,,