somewhat working ebpf bpftrace

BCC_TRACING.md

@@ -0,0 +1,298 @@
+# BCC-Style eBPF Tracing Implementation
+
+## Overview
+
+This implementation adds powerful BCC-style (Berkeley Packet Filter Compiler) tracing capabilities to the diagnostic agent, similar to the `trace.py` tool from the iovisor BCC project. Instead of just filtering events, this system actually counts and traces real system calls with detailed argument parsing.
+
+## Key Features
+
+### 1. Real System Call Tracing
+- **Actual event counting**: Unlike the previous implementation that just simulated events, this captures real system calls
+- **Argument extraction**: Extracts function arguments (arg1, arg2, etc.) and return values
+- **Multiple probe types**: Supports kprobes, kretprobes, tracepoints, and uprobes
+- **Filtering capabilities**: Filter by process name, PID, UID, argument values
+
+### 2. BCC-Style Syntax
+Supports familiar BCC trace.py syntax patterns:
+```bash
+# Simple syscall tracing
+"sys_open"                    # Trace open syscalls
+"sys_read (arg3 > 1024)"      # Trace reads >1024 bytes
+"r::sys_open"                 # Return probe on open
+
+# With format strings
+"sys_write \"wrote %d bytes\", arg3"
+"sys_open \"opening %s\", arg2@user"
+```
+
+### 3. Comprehensive Event Data
+Each trace captures:
+```json
+{
+  "timestamp": 1234567890,
+  "pid": 1234,
+  "tid": 1234,
+  "process_name": "nginx",
+  "function": "__x64_sys_openat",
+  "message": "opening file: /var/log/access.log",
+  "raw_args": {
+    "arg1": "3",
+    "arg2": "/var/log/access.log",
+    "arg3": "577"
+  }
+}
+```
+
+## Architecture
+
+### Core Components
+
+1. **BCCTraceManager** (`ebpf_trace_manager.go`)
+   - Main orchestrator for BCC-style tracing
+   - Generates bpftrace scripts dynamically
+   - Manages trace sessions and event collection
+
+2. **TraceSpec** - Trace specification format
+   ```go
+   type TraceSpec struct {
+       ProbeType    string            // "p", "r", "t", "u"
+       Target       string            // Function/syscall to trace
+       Format       string            // Output format string
+       Arguments    []string          // Arguments to extract
+       Filter       string            // Filter conditions
+       Duration     int               // Trace duration in seconds
+       ProcessName  string            // Process filter
+       PID          int               // Process ID filter
+       UID          int               // User ID filter
+   }
+   ```
+
+3. **EventScanner** (`ebpf_event_parser.go`)
+   - Parses bpftrace output in real-time
+   - Converts raw trace data to structured events
+   - Handles argument extraction and enrichment
+
+4. **TraceSpecBuilder** - Fluent API for building specs
+   ```go
+   spec := NewTraceSpecBuilder().
+       Kprobe("__x64_sys_write").
+       Format("write %d bytes to fd %d", "arg3", "arg1").
+       Filter("arg1 == 1").
+       Duration(30).
+       Build()
+   ```
+
+## Usage Examples
+
+### 1. Basic System Call Tracing
+
+```go
+// Trace file open operations
+spec := TraceSpec{
+    ProbeType: "p",
+    Target:    "__x64_sys_openat",
+    Format:    "opening file: %s",
+    Arguments: []string{"arg2@user"},
+    Duration:  30,
+}
+
+traceID, err := manager.StartTrace(spec)
+```
+
+### 2. Filtered Tracing
+
+```go
+// Trace only large reads
+spec := TraceSpec{
+    ProbeType: "p",
+    Target:    "__x64_sys_read",
+    Format:    "read %d bytes from fd %d",
+    Arguments: []string{"arg3", "arg1"},
+    Filter:    "arg3 > 1024",
+    Duration:  30,
+}
+```
+
+### 3. Process-Specific Tracing
+
+```go
+// Trace only nginx processes
+spec := TraceSpec{
+    ProbeType:   "p",
+    Target:      "__x64_sys_write",
+    ProcessName: "nginx",
+    Duration:    60,
+}
+```
+
+### 4. Return Value Tracing
+
+```go
+// Trace return values from file operations
+spec := TraceSpec{
+    ProbeType: "r",
+    Target:    "__x64_sys_openat",
+    Format:    "open returned: %d",
+    Arguments: []string{"retval"},
+    Duration:  30,
+}
+```
+
+## Integration with Agent
+
+### API Request Format
+The remote API can send trace specifications in the `ebpf_programs` field:
+
+```json
+{
+  "commands": [
+    {"id": "cmd1", "command": "ps aux"}
+  ],
+  "ebpf_programs": [
+    {
+      "name": "file_monitoring",
+      "type": "kprobe", 
+      "target": "sys_open",
+      "duration": 30,
+      "filters": {"process": "nginx"},
+      "description": "Monitor file access by nginx"
+    }
+  ]
+}
+```
+
+### Agent Response Format
+The agent returns detailed trace results:
+
+```json
+{
+  "name": "__x64_sys_openat",
+  "type": "bcc_trace",
+  "target": "__x64_sys_openat", 
+  "duration": 30,
+  "status": "completed",
+  "success": true,
+  "event_count": 45,
+  "events": [
+    {
+      "timestamp": 1234567890,
+      "pid": 1234,
+      "process_name": "nginx",
+      "function": "__x64_sys_openat",
+      "message": "opening file: /var/log/access.log",
+      "raw_args": {"arg1": "3", "arg2": "/var/log/access.log"}
+    }
+  ],
+  "statistics": {
+    "total_events": 45,
+    "events_per_second": 1.5,
+    "top_processes": [
+      {"process_name": "nginx", "event_count": 30},
+      {"process_name": "apache", "event_count": 15}
+    ]
+  }
+}
+```
+
+## Test Specifications
+
+The implementation includes test specifications for unit testing:
+
+- **test_sys_open**: File open operations
+- **test_sys_read**: Read operations with filters
+- **test_sys_write**: Write operations  
+- **test_process_creation**: Process execution
+- **test_kretprobe**: Return value tracing
+- **test_with_filter**: Filtered tracing
+
+## Running Tests
+
+```bash
+# Run all BCC tracing tests
+go test -v -run TestBCCTracing
+
+# Test trace manager capabilities
+go test -v -run TestTraceManagerCapabilities
+
+# Test syscall suggestions
+go test -v -run TestSyscallSuggestions
+
+# Run all tests
+go test -v
+```
+
+## Requirements
+
+### System Requirements
+- **Linux kernel 4.4+** with eBPF support
+- **bpftrace** installed (`apt install bpftrace`)
+- **Root privileges** for actual tracing
+
+### Checking Capabilities
+The trace manager automatically detects capabilities:
+
+```bash
+$ go test -run TestTraceManagerCapabilities
+🔧 Trace Manager Capabilities:
+   ✅ kernel_ebpf: Available
+   ✅ bpftrace: Available  
+   ❌ root_access: Not Available
+   ❌ debugfs_access: Not Available
+```
+
+## Advanced Features
+
+### 1. Syscall Suggestions
+The system can suggest appropriate syscalls based on issue descriptions:
+
+```go
+suggestions := SuggestSyscallTargets("file not found error")
+// Returns: ["test_sys_open", "test_sys_read", "test_sys_write", "test_sys_unlink"]
+```
+
+### 2. BCC-Style Parsing
+Parse BCC trace.py style specifications:
+
+```go
+parser := NewTraceSpecParser()
+spec, err := parser.ParseFromBCCStyle("sys_write (arg1 == 1) \"stdout: %d bytes\", arg3")
+```
+
+### 3. Event Filtering and Aggregation
+Post-processing capabilities for trace events:
+
+```go
+filter := &TraceEventFilter{
+    ProcessNames: []string{"nginx", "apache"},
+    MinTimestamp: startTime,
+}
+filteredEvents := filter.ApplyFilter(events)
+
+aggregator := NewTraceEventAggregator(events)
+topProcesses := aggregator.GetTopProcesses(5)
+eventRate := aggregator.GetEventRate()
+```
+
+## Performance Considerations
+
+- **Short durations**: Test specs use 5-second durations for quick testing
+- **Efficient parsing**: Event scanner processes bpftrace output in real-time
+- **Memory management**: Events are processed and aggregated efficiently
+- **Timeout handling**: Automatic cleanup of hanging trace sessions
+
+## Security Considerations
+
+- **Root privileges required**: eBPF tracing requires root access
+- **Resource limits**: Maximum trace duration of 10 minutes
+- **Process isolation**: Each trace runs in its own context
+- **Automatic cleanup**: Traces are automatically stopped and cleaned up
+
+## Future Enhancements
+
+1. **USDT probe support**: Add support for user-space tracing
+2. **BTF integration**: Use BPF Type Format for better type information  
+3. **Flame graph generation**: Generate performance flame graphs
+4. **Custom eBPF programs**: Allow uploading custom eBPF bytecode
+5. **Distributed tracing**: Correlation across multiple hosts
+
+This implementation provides a solid foundation for advanced system introspection and debugging, bringing the power of BCC-style tracing to the diagnostic agent.

agent.go

@@ -11,6 +11,9 @@ import (
 	"sync"
 	"time"
 
+	"nannyagentv2/internal/logging"
+	"nannyagentv2/internal/types"
+
 	"github.com/sashabaranov/go-openai"
 )
 
@@ -25,28 +28,6 @@ type DiagnosticResponse struct {
 	ConfidenceLevel float64  `json:"confidence_level"`
 }
 
-// EBPFRequest represents a request for eBPF program execution
-type EBPFRequest struct {
-	Name        string            `json:"name"`
-	Type        string            `json:"type"`
-	Target      string            `json:"target"`
-	Duration    int               `json:"duration"`
-	Filters     map[string]string `json:"filters,omitempty"`
-	Description string            `json:"description"`
-}
-
-// EBPFEnhancedDiagnosticResponse represents the enhanced diagnostic response with eBPF
-type EBPFEnhancedDiagnosticResponse struct {
-	ResponseType    string        `json:"response_type"`
-	Phase           string        `json:"phase"`
-	Analysis        string        `json:"analysis"`
-	Commands        []string      `json:"commands"`
-	EBPFPrograms    []EBPFRequest `json:"ebpf_programs"`
-	NextSteps       []string      `json:"next_steps"`
-	Reasoning       string        `json:"reasoning"`
-	ConfidenceLevel float64       `json:"confidence_level"`
-}
-
 // ResolutionResponse represents the resolution phase response from AI
 type ResolutionResponse struct {
 	ResponseType   string `json:"response_type"`
@@ -93,6 +74,7 @@ type LinuxDiagnosticAgent struct {
 	episodeID   string           // TensorZero episode ID for conversation continuity
 	ebpfManager *BCCTraceManager // BCC-style eBPF tracing capabilities
 	config      *AgentConfig     // Configuration for concurrent execution
+	authManager interface{}      // Authentication manager for TensorZero requests
 }
 
 // NewLinuxDiagnosticAgent creates a new diagnostic agent
@@ -100,14 +82,14 @@ func NewLinuxDiagnosticAgent() *LinuxDiagnosticAgent {
 	// Get Supabase project URL for TensorZero proxy
 	supabaseURL := os.Getenv("SUPABASE_PROJECT_URL")
 	if supabaseURL == "" {
-		fmt.Printf("Warning: SUPABASE_PROJECT_URL not set, TensorZero integration will not work\n")
+		logging.Warning("SUPABASE_PROJECT_URL not set, TensorZero integration will not work")
 		supabaseURL = "https://gpqzsricripnvbrpsyws.supabase.co" // fallback
 	}
 
 	model := os.Getenv("NANNYAPI_MODEL")
 	if model == "" {
 		model = "tensorzero::function_name::diagnose_and_heal"
-		fmt.Printf("Warning: Using default model '%s'. Set NANNYAPI_MODEL environment variable for your specific function.\n", model)
+		logging.Warning("Using default model '%s'. Set NANNYAPI_MODEL environment variable for your specific function", model)
 	}
 
 	// Note: We don't use the OpenAI client anymore, we use direct HTTP to Supabase proxy
@@ -124,10 +106,40 @@ func NewLinuxDiagnosticAgent() *LinuxDiagnosticAgent {
 	return agent
 }
 
+// NewLinuxDiagnosticAgentWithAuth creates a new diagnostic agent with authentication
+func NewLinuxDiagnosticAgentWithAuth(authManager interface{}) *LinuxDiagnosticAgent {
+	// Get Supabase project URL for TensorZero proxy
+	supabaseURL := os.Getenv("SUPABASE_PROJECT_URL")
+	if supabaseURL == "" {
+		logging.Warning("SUPABASE_PROJECT_URL not set, TensorZero integration will not work")
+		supabaseURL = "https://gpqzsricripnvbrpsyws.supabase.co" // fallback
+	}
+
+	model := os.Getenv("NANNYAPI_MODEL")
+	if model == "" {
+		model = "tensorzero::function_name::diagnose_and_heal"
+		logging.Warning("Using default model '%s'. Set NANNYAPI_MODEL environment variable for your specific function", model)
+	}
+
+	// Note: We don't use the OpenAI client anymore, we use direct HTTP to Supabase proxy
+	agent := &LinuxDiagnosticAgent{
+		client:      nil, // Not used anymore
+		model:       model,
+		executor:    NewCommandExecutor(10 * time.Second), // 10 second timeout for commands
+		config:      DefaultAgentConfig(),                 // Default concurrent execution config
+		authManager: authManager,                          // Store auth manager for TensorZero requests
+	}
+
+	// Initialize BCC-style eBPF capabilities
+	agent.ebpfManager = NewBCCTraceManager()
+
+	return agent
+}
+
 // DiagnoseIssue starts the diagnostic process for a given issue
 func (a *LinuxDiagnosticAgent) DiagnoseIssue(issue string) error {
-	fmt.Printf("Diagnosing issue: %s\n", issue)
-	fmt.Println("Gathering system information...")
+	logging.Info("Diagnosing issue: %s", issue)
+	logging.Info("Gathering system information...")
 
 	// Gather system information
 	systemInfo := GatherSystemInfo()
@@ -155,25 +167,27 @@ func (a *LinuxDiagnosticAgent) DiagnoseIssue(issue string) error {
 		}
 
 		content := response.Choices[0].Message.Content
-		fmt.Printf("\nAI Response:\n%s\n", content)
+		logging.Debug("AI Response: %s", content)
 
 		// Parse the response to determine next action
-		var diagnosticResp EBPFEnhancedDiagnosticResponse
+		var diagnosticResp types.EBPFEnhancedDiagnosticResponse
 		var resolutionResp ResolutionResponse
 
 		// Try to parse as diagnostic response first (with eBPF support)
+		logging.Debug("Attempting to parse response as diagnostic...")
 		if err := json.Unmarshal([]byte(content), &diagnosticResp); err == nil && diagnosticResp.ResponseType == "diagnostic" {
+			logging.Debug("Successfully parsed as diagnostic response with %d commands", len(diagnosticResp.Commands))
 			// Handle diagnostic phase
-			fmt.Printf("\nReasoning: %s\n", diagnosticResp.Reasoning)
+			logging.Debug("Reasoning: %s", diagnosticResp.Reasoning)
 
 			// Execute commands and collect results
 			commandResults := make([]CommandResult, 0, len(diagnosticResp.Commands))
 			if len(diagnosticResp.Commands) > 0 {
-				fmt.Printf("🔧 Executing diagnostic commands...\n")
+				logging.Info("Executing %d diagnostic commands", len(diagnosticResp.Commands))
 				for i, cmdStr := range diagnosticResp.Commands {
-					// Convert string to Command struct
+					// Convert string command to Command struct (auto-generate ID and description)
 					cmd := Command{
-						ID:          fmt.Sprintf("cmd_%d", i),
+						ID:          fmt.Sprintf("cmd_%d", i+1),
 						Command:     cmdStr,
 						Description: fmt.Sprintf("Diagnostic command: %s", cmdStr),
 					}
@@ -181,7 +195,7 @@ func (a *LinuxDiagnosticAgent) DiagnoseIssue(issue string) error {
 					commandResults = append(commandResults, result)
 
 					if result.ExitCode != 0 {
-						fmt.Printf("❌ Command '%s' failed with exit code %d\n", cmd.ID, result.ExitCode)
+						logging.Warning("Command '%s' failed with exit code %d", cmd.ID, result.ExitCode)
 					}
 				}
 			}
@@ -189,7 +203,7 @@ func (a *LinuxDiagnosticAgent) DiagnoseIssue(issue string) error {
 			// Execute eBPF programs if present - support both old and new formats
 			var ebpfResults []map[string]interface{}
 			if len(diagnosticResp.EBPFPrograms) > 0 {
-				fmt.Printf("🔬 AI requested %d eBPF traces for enhanced diagnostics\n", len(diagnosticResp.EBPFPrograms))
+				logging.Info("AI requested %d eBPF traces for enhanced diagnostics", len(diagnosticResp.EBPFPrograms))
 
 				// Convert EBPFPrograms to TraceSpecs and execute concurrently
 				traceSpecs := a.convertEBPFProgramsToTraceSpecs(diagnosticResp.EBPFPrograms)
@@ -210,12 +224,17 @@ func (a *LinuxDiagnosticAgent) DiagnoseIssue(issue string) error {
 				// Extract evidence summary for TensorZero
 				evidenceSummary := make([]string, 0)
 				for _, result := range ebpfResults {
-					name := result["name"]
-					eventCount := result["data_points"]
-					description := result["description"]
-					status := result["status"]
+					target := result["target"]
+					eventCount := result["event_count"]
+					summary := result["summary"]
+					success := result["success"]
 
-					summaryStr := fmt.Sprintf("%s: %v events (%s) - %s", name, eventCount, status, description)
+					status := "failed"
+					if success == true {
+						status = "success"
+					}
+
+					summaryStr := fmt.Sprintf("%s: %v events (%s) - %s", target, eventCount, status, summary)
 					evidenceSummary = append(evidenceSummary, summaryStr)
 				}
 				allResults["ebpf_evidence_summary"] = evidenceSummary
@@ -237,20 +256,22 @@ func (a *LinuxDiagnosticAgent) DiagnoseIssue(issue string) error {
 			})
 
 			continue
+		} else {
+			logging.Debug("Failed to parse as diagnostic. Error: %v, ResponseType: '%s'", err, diagnosticResp.ResponseType)
 		}
 
 		// Try to parse as resolution response
 		if err := json.Unmarshal([]byte(content), &resolutionResp); err == nil && resolutionResp.ResponseType == "resolution" {
 			// Handle resolution phase
-			fmt.Printf("\n=== DIAGNOSIS COMPLETE ===\n")
-			fmt.Printf("Root Cause: %s\n", resolutionResp.RootCause)
-			fmt.Printf("Resolution Plan: %s\n", resolutionResp.ResolutionPlan)
-			fmt.Printf("Confidence: %s\n", resolutionResp.Confidence)
+			logging.Info("=== DIAGNOSIS COMPLETE ===")
+			logging.Info("Root Cause: %s", resolutionResp.RootCause)
+			logging.Info("Resolution Plan: %s", resolutionResp.ResolutionPlan)
+			logging.Info("Confidence: %s", resolutionResp.Confidence)
 			break
 		}
 
 		// If we can't parse the response, treat it as an error or unexpected format
-		fmt.Printf("Unexpected response format or error from AI:\n%s\n", content)
+		logging.Error("Unexpected response format or error from AI: %s", content)
 		break
 	}
 
@@ -296,8 +317,9 @@ func (a *LinuxDiagnosticAgent) sendRequestWithEpisode(messages []openai.ChatComp
 		return nil, fmt.Errorf("SUPABASE_PROJECT_URL not set")
 	}
 
-	// Create HTTP request to TensorZero proxy
-	endpoint := fmt.Sprintf("%s/functions/v1/tensorzero-proxy", supabaseURL)
+	// Create HTTP request to TensorZero proxy (includes OpenAI-compatible path)
+	endpoint := fmt.Sprintf("%s/functions/v1/tensorzero-proxy/openai/v1/chat/completions", supabaseURL)
+	logging.Debug("Calling TensorZero proxy at: %s", endpoint)
 	req, err := http.NewRequest("POST", endpoint, bytes.NewBuffer(requestBody))
 	if err != nil {
 		return nil, fmt.Errorf("failed to create request: %w", err)
@@ -307,7 +329,17 @@ func (a *LinuxDiagnosticAgent) sendRequestWithEpisode(messages []openai.ChatComp
 	req.Header.Set("Content-Type", "application/json")
 	req.Header.Set("Accept", "application/json")
 
-	// Note: No authentication needed for TensorZero proxy based on the existing pattern
+	// Add authentication if auth manager is available (same pattern as investigation_server.go)
+	if a.authManager != nil {
+		// The authManager should be *auth.AuthManager, so let's use the exact same pattern
+		if authMgr, ok := a.authManager.(interface {
+			LoadToken() (*types.AuthToken, error)
+		}); ok {
+			if authToken, err := authMgr.LoadToken(); err == nil && authToken != nil {
+				req.Header.Set("Authorization", fmt.Sprintf("Bearer %s", authToken.AccessToken))
+			}
+		}
+	}
 
 	// Send request
 	client := &http.Client{Timeout: 30 * time.Second}
@@ -372,7 +404,7 @@ func (a *LinuxDiagnosticAgent) sendRequestWithEpisode(messages []openai.ChatComp
 }
 
 // convertEBPFProgramsToTraceSpecs converts old EBPFProgram format to new TraceSpec format
-func (a *LinuxDiagnosticAgent) convertEBPFProgramsToTraceSpecs(ebpfPrograms []EBPFRequest) []TraceSpec {
+func (a *LinuxDiagnosticAgent) convertEBPFProgramsToTraceSpecs(ebpfPrograms []types.EBPFRequest) []TraceSpec {
 	var traceSpecs []TraceSpec
 
 	for _, prog := range ebpfPrograms {
@@ -384,7 +416,7 @@ func (a *LinuxDiagnosticAgent) convertEBPFProgramsToTraceSpecs(ebpfPrograms []EB
 }
 
 // convertToTraceSpec converts an EBPFRequest to a TraceSpec for BCC-style tracing
-func (a *LinuxDiagnosticAgent) convertToTraceSpec(prog EBPFRequest) TraceSpec {
+func (a *LinuxDiagnosticAgent) convertToTraceSpec(prog types.EBPFRequest) TraceSpec {
 	// Determine probe type based on target and type
 	probeType := "p" // default to kprobe
 	target := prog.Target
@@ -421,6 +453,7 @@ func (a *LinuxDiagnosticAgent) convertToTraceSpec(prog EBPFRequest) TraceSpec {
 		Format:    prog.Description, // Use description as format
 		Arguments: []string{},       // Start with no arguments for compatibility
 		Duration:  duration,
+		UID:       -1, // No UID filter (don't default to 0 which means root only)
 	}
 }
 
@@ -430,7 +463,7 @@ func (a *LinuxDiagnosticAgent) executeBCCTracesConcurrently(traceSpecs []TraceSp
 		return []map[string]interface{}{}
 	}
 
-	fmt.Printf("🚀 Executing %d BCC traces with max %d concurrent tasks\n", len(traceSpecs), a.config.MaxConcurrentTasks)
+	logging.Info("Executing %d BCC traces with max %d concurrent tasks", len(traceSpecs), a.config.MaxConcurrentTasks)
 
 	// Channel to limit concurrent goroutines
 	semaphore := make(chan struct{}, a.config.MaxConcurrentTasks)
@@ -465,7 +498,7 @@ func (a *LinuxDiagnosticAgent) executeBCCTracesConcurrently(traceSpecs []TraceSp
 	}
 
 	if a.config.CollectiveResults {
-		fmt.Printf("✅ All %d BCC traces completed. Sending collective results to API layer.\n", len(allResults))
+		logging.Debug("All %d BCC traces completed. Sending collective results to API layer", len(allResults))
 	}
 
 	return allResults
@@ -482,18 +515,18 @@ func (a *LinuxDiagnosticAgent) executeSingleBCCTrace(index int, spec TraceSpec)
 		"start_time": time.Now().Format(time.RFC3339),
 	}
 
-	fmt.Printf("🔍 [Task %d] Starting BCC trace: %s (type: %s)\n", index, spec.Target, spec.ProbeType)
+	logging.Debug("[Task %d] Starting BCC trace: %s (type: %s)", index, spec.Target, spec.ProbeType)
 
 	// Start the trace
 	traceID, err := a.ebpfManager.StartTrace(spec)
 	if err != nil {
 		result["error"] = fmt.Sprintf("Failed to start trace: %v", err)
-		fmt.Printf("❌ [Task %d] Failed to start trace %s: %v\n", index, spec.Target, err)
+		logging.Error("[Task %d] Failed to start trace %s: %v", index, spec.Target, err)
 		return result
 	}
 
 	result["trace_id"] = traceID
-	fmt.Printf("🚀 [Task %d] Trace %s started with ID: %s\n", index, spec.Target, traceID)
+	logging.Debug("[Task %d] Trace %s started with ID: %s", index, spec.Target, traceID)
 
 	// Wait for the trace duration
 	time.Sleep(time.Duration(spec.Duration) * time.Second)
@@ -504,7 +537,7 @@ func (a *LinuxDiagnosticAgent) executeSingleBCCTrace(index int, spec TraceSpec)
 		// Try to stop the trace if it's still running
 		a.ebpfManager.StopTrace(traceID)
 		result["error"] = fmt.Sprintf("Failed to get trace results: %v", err)
-		fmt.Printf("❌ [Task %d] Failed to get results for trace %s: %v\n", index, spec.Target, err)
+		logging.Error("[Task %d] Failed to get results for trace %s: %v", index, spec.Target, err)
 		return result
 	}
 
@@ -551,7 +584,7 @@ func (a *LinuxDiagnosticAgent) executeSingleBCCTrace(index int, spec TraceSpec)
 		result["top_processes"] = topProcesses
 	}
 
-	fmt.Printf("✅ [Task %d] Trace %s completed: %d events (%.2f events/sec)\n",
+	logging.Debug("[Task %d] Trace %s completed: %d events (%.2f events/sec)",
 		index, spec.Target, traceResult.EventCount, traceResult.Statistics.EventsPerSecond)
 
 	return result

ebpf_event_parser.go

@@ -0,0 +1,343 @@
+package main
+
+import (
+	"bufio"
+	"io"
+	"regexp"
+	"strconv"
+	"strings"
+	"time"
+)
+
+// EventScanner parses bpftrace output and converts it to TraceEvent structs
+type EventScanner struct {
+	scanner   *bufio.Scanner
+	lastEvent *TraceEvent
+	lineRegex *regexp.Regexp
+}
+
+// NewEventScanner creates a new event scanner for parsing bpftrace output
+func NewEventScanner(reader io.Reader) *EventScanner {
+	// Regex pattern to match our trace output format:
+	// TRACE|timestamp|pid|tid|comm|function|message
+	pattern := `^TRACE\|(\d+)\|(\d+)\|(\d+)\|([^|]+)\|([^|]+)\|(.*)$`
+	regex, _ := regexp.Compile(pattern)
+
+	return &EventScanner{
+		scanner:   bufio.NewScanner(reader),
+		lineRegex: regex,
+	}
+}
+
+// Scan advances the scanner to the next event
+func (es *EventScanner) Scan() bool {
+	for es.scanner.Scan() {
+		line := strings.TrimSpace(es.scanner.Text())
+
+		// Skip empty lines and non-trace lines
+		if line == "" || !strings.HasPrefix(line, "TRACE|") {
+			continue
+		}
+
+		// Parse the trace line
+		if event := es.parseLine(line); event != nil {
+			es.lastEvent = event
+			return true
+		}
+	}
+
+	return false
+}
+
+// Event returns the most recently parsed event
+func (es *EventScanner) Event() *TraceEvent {
+	return es.lastEvent
+}
+
+// Error returns any scanning error
+func (es *EventScanner) Error() error {
+	return es.scanner.Err()
+}
+
+// parseLine parses a single trace line into a TraceEvent
+func (es *EventScanner) parseLine(line string) *TraceEvent {
+	matches := es.lineRegex.FindStringSubmatch(line)
+	if len(matches) != 7 {
+		return nil
+	}
+
+	// Parse timestamp (nanoseconds)
+	timestamp, err := strconv.ParseInt(matches[1], 10, 64)
+	if err != nil {
+		return nil
+	}
+
+	// Parse PID
+	pid, err := strconv.Atoi(matches[2])
+	if err != nil {
+		return nil
+	}
+
+	// Parse TID
+	tid, err := strconv.Atoi(matches[3])
+	if err != nil {
+		return nil
+	}
+
+	// Extract process name, function, and message
+	processName := strings.TrimSpace(matches[4])
+	function := strings.TrimSpace(matches[5])
+	message := strings.TrimSpace(matches[6])
+
+	event := &TraceEvent{
+		Timestamp:   timestamp,
+		PID:         pid,
+		TID:         tid,
+		ProcessName: processName,
+		Function:    function,
+		Message:     message,
+		RawArgs:     make(map[string]string),
+	}
+
+	// Try to extract additional information from the message
+	es.enrichEvent(event, message)
+
+	return event
+}
+
+// enrichEvent extracts additional information from the message
+func (es *EventScanner) enrichEvent(event *TraceEvent, message string) {
+	// Parse common patterns in messages to extract arguments
+	// This is a simplified version - in a real implementation you'd want more sophisticated parsing
+
+	// Look for patterns like "arg1=value, arg2=value"
+	argPattern := regexp.MustCompile(`(\w+)=([^,\s]+)`)
+	matches := argPattern.FindAllStringSubmatch(message, -1)
+
+	for _, match := range matches {
+		if len(match) == 3 {
+			event.RawArgs[match[1]] = match[2]
+		}
+	}
+
+	// Look for numeric patterns that might be syscall arguments
+	numberPattern := regexp.MustCompile(`\b(\d+)\b`)
+	numbers := numberPattern.FindAllString(message, -1)
+
+	for i, num := range numbers {
+		argName := "arg" + strconv.Itoa(i+1)
+		event.RawArgs[argName] = num
+	}
+}
+
+// TraceEventFilter provides filtering capabilities for trace events
+type TraceEventFilter struct {
+	MinTimestamp  int64
+	MaxTimestamp  int64
+	ProcessNames  []string
+	PIDs          []int
+	UIDs          []int
+	Functions     []string
+	MessageFilter string
+}
+
+// ApplyFilter applies filters to a slice of events
+func (filter *TraceEventFilter) ApplyFilter(events []TraceEvent) []TraceEvent {
+	if filter == nil {
+		return events
+	}
+
+	var filtered []TraceEvent
+
+	for _, event := range events {
+		if filter.matchesEvent(&event) {
+			filtered = append(filtered, event)
+		}
+	}
+
+	return filtered
+}
+
+// matchesEvent checks if an event matches the filter criteria
+func (filter *TraceEventFilter) matchesEvent(event *TraceEvent) bool {
+	// Check timestamp range
+	if filter.MinTimestamp > 0 && event.Timestamp < filter.MinTimestamp {
+		return false
+	}
+	if filter.MaxTimestamp > 0 && event.Timestamp > filter.MaxTimestamp {
+		return false
+	}
+
+	// Check process names
+	if len(filter.ProcessNames) > 0 {
+		found := false
+		for _, name := range filter.ProcessNames {
+			if strings.Contains(event.ProcessName, name) {
+				found = true
+				break
+			}
+		}
+		if !found {
+			return false
+		}
+	}
+
+	// Check PIDs
+	if len(filter.PIDs) > 0 {
+		found := false
+		for _, pid := range filter.PIDs {
+			if event.PID == pid {
+				found = true
+				break
+			}
+		}
+		if !found {
+			return false
+		}
+	}
+
+	// Check UIDs
+	if len(filter.UIDs) > 0 {
+		found := false
+		for _, uid := range filter.UIDs {
+			if event.UID == uid {
+				found = true
+				break
+			}
+		}
+		if !found {
+			return false
+		}
+	}
+
+	// Check functions
+	if len(filter.Functions) > 0 {
+		found := false
+		for _, function := range filter.Functions {
+			if strings.Contains(event.Function, function) {
+				found = true
+				break
+			}
+		}
+		if !found {
+			return false
+		}
+	}
+
+	// Check message filter
+	if filter.MessageFilter != "" {
+		if !strings.Contains(event.Message, filter.MessageFilter) {
+			return false
+		}
+	}
+
+	return true
+}
+
+// TraceEventAggregator provides aggregation capabilities for trace events
+type TraceEventAggregator struct {
+	events []TraceEvent
+}
+
+// NewTraceEventAggregator creates a new event aggregator
+func NewTraceEventAggregator(events []TraceEvent) *TraceEventAggregator {
+	return &TraceEventAggregator{
+		events: events,
+	}
+}
+
+// CountByProcess returns event counts grouped by process
+func (agg *TraceEventAggregator) CountByProcess() map[string]int {
+	counts := make(map[string]int)
+	for _, event := range agg.events {
+		counts[event.ProcessName]++
+	}
+	return counts
+}
+
+// CountByFunction returns event counts grouped by function
+func (agg *TraceEventAggregator) CountByFunction() map[string]int {
+	counts := make(map[string]int)
+	for _, event := range agg.events {
+		counts[event.Function]++
+	}
+	return counts
+}
+
+// CountByPID returns event counts grouped by PID
+func (agg *TraceEventAggregator) CountByPID() map[int]int {
+	counts := make(map[int]int)
+	for _, event := range agg.events {
+		counts[event.PID]++
+	}
+	return counts
+}
+
+// GetTimeRange returns the time range of events
+func (agg *TraceEventAggregator) GetTimeRange() (int64, int64) {
+	if len(agg.events) == 0 {
+		return 0, 0
+	}
+
+	minTime := agg.events[0].Timestamp
+	maxTime := agg.events[0].Timestamp
+
+	for _, event := range agg.events {
+		if event.Timestamp < minTime {
+			minTime = event.Timestamp
+		}
+		if event.Timestamp > maxTime {
+			maxTime = event.Timestamp
+		}
+	}
+
+	return minTime, maxTime
+}
+
+// GetEventRate calculates events per second
+func (agg *TraceEventAggregator) GetEventRate() float64 {
+	if len(agg.events) < 2 {
+		return 0
+	}
+
+	minTime, maxTime := agg.GetTimeRange()
+	durationNs := maxTime - minTime
+	durationSeconds := float64(durationNs) / float64(time.Second)
+
+	if durationSeconds == 0 {
+		return 0
+	}
+
+	return float64(len(agg.events)) / durationSeconds
+}
+
+// GetTopProcesses returns the most active processes
+func (agg *TraceEventAggregator) GetTopProcesses(limit int) []ProcessStat {
+	counts := agg.CountByProcess()
+	total := len(agg.events)
+
+	var stats []ProcessStat
+	for processName, count := range counts {
+		percentage := float64(count) / float64(total) * 100
+		stats = append(stats, ProcessStat{
+			ProcessName: processName,
+			EventCount:  count,
+			Percentage:  percentage,
+		})
+	}
+
+	// Simple sorting by event count (bubble sort for simplicity)
+	for i := 0; i < len(stats); i++ {
+		for j := i + 1; j < len(stats); j++ {
+			if stats[j].EventCount > stats[i].EventCount {
+				stats[i], stats[j] = stats[j], stats[i]
+			}
+		}
+	}
+
+	if limit > 0 && limit < len(stats) {
+		stats = stats[:limit]
+	}
+
+	return stats
+}

ebpf_trace_manager.go

@@ -0,0 +1,587 @@
+package main
+
+import (
+	"context"
+	"fmt"
+	"io"
+	"os"
+	"os/exec"
+	"strings"
+	"sync"
+	"time"
+
+	"nannyagentv2/internal/logging"
+)
+
+// TraceSpec represents a trace specification similar to BCC trace.py
+type TraceSpec struct {
+	// Probe type: "p" (kprobe), "r" (kretprobe), "t" (tracepoint), "u" (uprobe)
+	ProbeType string `json:"probe_type"`
+
+	// Target function/syscall/tracepoint
+	Target string `json:"target"`
+
+	// Library for userspace probes (empty for kernel)
+	Library string `json:"library,omitempty"`
+
+	// Format string for output (e.g., "read %d bytes", arg3)
+	Format string `json:"format"`
+
+	// Arguments to extract (e.g., ["arg1", "arg2", "retval"])
+	Arguments []string `json:"arguments"`
+
+	// Filter condition (e.g., "arg3 > 20000")
+	Filter string `json:"filter,omitempty"`
+
+	// Duration in seconds
+	Duration int `json:"duration"`
+
+	// Process ID filter (optional)
+	PID int `json:"pid,omitempty"`
+
+	// Thread ID filter (optional)
+	TID int `json:"tid,omitempty"`
+
+	// UID filter (optional)
+	UID int `json:"uid,omitempty"`
+
+	// Process name filter (optional)
+	ProcessName string `json:"process_name,omitempty"`
+}
+
+// TraceEvent represents a captured event from eBPF
+type TraceEvent struct {
+	Timestamp   int64             `json:"timestamp"`
+	PID         int               `json:"pid"`
+	TID         int               `json:"tid"`
+	UID         int               `json:"uid"`
+	ProcessName string            `json:"process_name"`
+	Function    string            `json:"function"`
+	Message     string            `json:"message"`
+	RawArgs     map[string]string `json:"raw_args"`
+	CPU         int               `json:"cpu,omitempty"`
+}
+
+// TraceResult represents the results of a tracing session
+type TraceResult struct {
+	TraceID    string       `json:"trace_id"`
+	Spec       TraceSpec    `json:"spec"`
+	Events     []TraceEvent `json:"events"`
+	EventCount int          `json:"event_count"`
+	StartTime  time.Time    `json:"start_time"`
+	EndTime    time.Time    `json:"end_time"`
+	Summary    string       `json:"summary"`
+	Statistics TraceStats   `json:"statistics"`
+}
+
+// TraceStats provides statistics about the trace
+type TraceStats struct {
+	TotalEvents     int            `json:"total_events"`
+	EventsByProcess map[string]int `json:"events_by_process"`
+	EventsByUID     map[int]int    `json:"events_by_uid"`
+	EventsPerSecond float64        `json:"events_per_second"`
+	TopProcesses    []ProcessStat  `json:"top_processes"`
+}
+
+// ProcessStat represents statistics for a process
+type ProcessStat struct {
+	ProcessName string  `json:"process_name"`
+	PID         int     `json:"pid"`
+	EventCount  int     `json:"event_count"`
+	Percentage  float64 `json:"percentage"`
+}
+
+// BCCTraceManager implements advanced eBPF tracing similar to BCC trace.py
+type BCCTraceManager struct {
+	traces       map[string]*RunningTrace
+	tracesLock   sync.RWMutex
+	traceCounter int
+	capabilities map[string]bool
+}
+
+// RunningTrace represents an active trace session
+type RunningTrace struct {
+	ID        string
+	Spec      TraceSpec
+	Process   *exec.Cmd
+	Events    []TraceEvent
+	StartTime time.Time
+	Cancel    context.CancelFunc
+	Context   context.Context
+	Done      chan struct{} // Signal when trace monitoring is complete
+}
+
+// NewBCCTraceManager creates a new BCC-style trace manager
+func NewBCCTraceManager() *BCCTraceManager {
+	manager := &BCCTraceManager{
+		traces:       make(map[string]*RunningTrace),
+		capabilities: make(map[string]bool),
+	}
+
+	manager.testCapabilities()
+	return manager
+}
+
+// testCapabilities checks what tracing capabilities are available
+func (tm *BCCTraceManager) testCapabilities() {
+	// Test if bpftrace is available
+	if _, err := exec.LookPath("bpftrace"); err == nil {
+		tm.capabilities["bpftrace"] = true
+	} else {
+		tm.capabilities["bpftrace"] = false
+	}
+
+	// Test if perf is available for fallback
+	if _, err := exec.LookPath("perf"); err == nil {
+		tm.capabilities["perf"] = true
+	} else {
+		tm.capabilities["perf"] = false
+	}
+
+	// Test root privileges (required for eBPF)
+	tm.capabilities["root_access"] = os.Geteuid() == 0
+
+	// Test kernel version
+	cmd := exec.Command("uname", "-r")
+	output, err := cmd.Output()
+	if err == nil {
+		version := strings.TrimSpace(string(output))
+		// eBPF requires kernel 4.4+
+		tm.capabilities["kernel_ebpf"] = !strings.HasPrefix(version, "3.")
+	} else {
+		tm.capabilities["kernel_ebpf"] = false
+	}
+
+	// Test if we can access debugfs
+	if _, err := os.Stat("/sys/kernel/debug/tracing/available_events"); err == nil {
+		tm.capabilities["debugfs_access"] = true
+	} else {
+		tm.capabilities["debugfs_access"] = false
+	}
+
+	logging.Debug("BCC Trace capabilities: %+v", tm.capabilities)
+}
+
+// GetCapabilities returns available tracing capabilities
+func (tm *BCCTraceManager) GetCapabilities() map[string]bool {
+	tm.tracesLock.RLock()
+	defer tm.tracesLock.RUnlock()
+
+	caps := make(map[string]bool)
+	for k, v := range tm.capabilities {
+		caps[k] = v
+	}
+	return caps
+}
+
+// StartTrace starts a new trace session based on the specification
+func (tm *BCCTraceManager) StartTrace(spec TraceSpec) (string, error) {
+	if !tm.capabilities["bpftrace"] {
+		return "", fmt.Errorf("bpftrace not available - install bpftrace package")
+	}
+
+	if !tm.capabilities["root_access"] {
+		return "", fmt.Errorf("root access required for eBPF tracing")
+	}
+
+	if !tm.capabilities["kernel_ebpf"] {
+		return "", fmt.Errorf("kernel version does not support eBPF")
+	}
+
+	tm.tracesLock.Lock()
+	defer tm.tracesLock.Unlock()
+
+	// Generate trace ID
+	tm.traceCounter++
+	traceID := fmt.Sprintf("trace_%d", tm.traceCounter)
+
+	// Generate bpftrace script
+	script, err := tm.generateBpftraceScript(spec)
+	if err != nil {
+		return "", fmt.Errorf("failed to generate bpftrace script: %w", err)
+	}
+
+	// Debug: log the generated script
+	logging.Debug("Generated bpftrace script for %s:\n%s", spec.Target, script)
+
+	// Create context with timeout
+	ctx, cancel := context.WithTimeout(context.Background(), time.Duration(spec.Duration)*time.Second)
+
+	// Start bpftrace process
+	cmd := exec.CommandContext(ctx, "bpftrace", "-e", script)
+
+	// Create stdout pipe BEFORE starting
+	stdout, err := cmd.StdoutPipe()
+	if err != nil {
+		cancel()
+		return "", fmt.Errorf("failed to create stdout pipe: %w", err)
+	}
+
+	trace := &RunningTrace{
+		ID:        traceID,
+		Spec:      spec,
+		Process:   cmd,
+		Events:    []TraceEvent{},
+		StartTime: time.Now(),
+		Cancel:    cancel,
+		Context:   ctx,
+		Done:      make(chan struct{}), // Initialize completion signal
+	}
+
+	// Start the trace
+	if err := cmd.Start(); err != nil {
+		cancel()
+		return "", fmt.Errorf("failed to start bpftrace: %w", err)
+	}
+
+	tm.traces[traceID] = trace
+
+	// Monitor the trace in a goroutine
+	go tm.monitorTrace(traceID, stdout)
+
+	logging.Debug("Started BCC-style trace %s for target %s", traceID, spec.Target)
+	return traceID, nil
+} // generateBpftraceScript generates a bpftrace script based on the trace specification
+func (tm *BCCTraceManager) generateBpftraceScript(spec TraceSpec) (string, error) {
+	var script strings.Builder
+
+	// Build probe specification
+	var probe string
+	switch spec.ProbeType {
+	case "p", "": // kprobe (default)
+		if strings.HasPrefix(spec.Target, "sys_") || strings.HasPrefix(spec.Target, "__x64_sys_") {
+			probe = fmt.Sprintf("kprobe:%s", spec.Target)
+		} else {
+			probe = fmt.Sprintf("kprobe:%s", spec.Target)
+		}
+	case "r": // kretprobe
+		if strings.HasPrefix(spec.Target, "sys_") || strings.HasPrefix(spec.Target, "__x64_sys_") {
+			probe = fmt.Sprintf("kretprobe:%s", spec.Target)
+		} else {
+			probe = fmt.Sprintf("kretprobe:%s", spec.Target)
+		}
+	case "t": // tracepoint
+		// If target already includes tracepoint prefix, use as-is
+		if strings.HasPrefix(spec.Target, "tracepoint:") {
+			probe = spec.Target
+		} else {
+			probe = fmt.Sprintf("tracepoint:%s", spec.Target)
+		}
+	case "u": // uprobe
+		if spec.Library == "" {
+			return "", fmt.Errorf("library required for uprobe")
+		}
+		probe = fmt.Sprintf("uprobe:%s:%s", spec.Library, spec.Target)
+	default:
+		return "", fmt.Errorf("unsupported probe type: %s", spec.ProbeType)
+	}
+
+	// Add BEGIN block
+	script.WriteString("BEGIN {\n")
+	script.WriteString(fmt.Sprintf("  printf(\"Starting trace for %s...\\n\");\n", spec.Target))
+	script.WriteString("}\n\n")
+
+	// Build the main probe
+	script.WriteString(fmt.Sprintf("%s {\n", probe))
+
+	// Add filters if specified
+	if tm.needsFiltering(spec) {
+		script.WriteString("  if (")
+		filters := tm.buildFilters(spec)
+		script.WriteString(strings.Join(filters, " && "))
+		script.WriteString(") {\n")
+	}
+
+	// Build output format
+	outputFormat := tm.buildOutputFormat(spec)
+	script.WriteString(fmt.Sprintf("    printf(\"%s\\n\"", outputFormat))
+
+	// Add arguments
+	args := tm.buildArgumentList(spec)
+	if len(args) > 0 {
+		script.WriteString(", ")
+		script.WriteString(strings.Join(args, ", "))
+	}
+
+	script.WriteString(");\n")
+
+	// Close filter if block
+	if tm.needsFiltering(spec) {
+		script.WriteString("  }\n")
+	}
+
+	script.WriteString("}\n\n")
+
+	// Add END block
+	script.WriteString("END {\n")
+	script.WriteString(fmt.Sprintf("  printf(\"Trace completed for %s\\n\");\n", spec.Target))
+	script.WriteString("}\n")
+
+	return script.String(), nil
+}
+
+// needsFiltering checks if any filters are needed
+func (tm *BCCTraceManager) needsFiltering(spec TraceSpec) bool {
+	return spec.PID != 0 || spec.TID != 0 || spec.UID != -1 ||
+		spec.ProcessName != "" || spec.Filter != ""
+}
+
+// buildFilters builds the filter conditions
+func (tm *BCCTraceManager) buildFilters(spec TraceSpec) []string {
+	var filters []string
+
+	if spec.PID != 0 {
+		filters = append(filters, fmt.Sprintf("pid == %d", spec.PID))
+	}
+
+	if spec.TID != 0 {
+		filters = append(filters, fmt.Sprintf("tid == %d", spec.TID))
+	}
+
+	if spec.UID != -1 {
+		filters = append(filters, fmt.Sprintf("uid == %d", spec.UID))
+	}
+
+	if spec.ProcessName != "" {
+		filters = append(filters, fmt.Sprintf("strncmp(comm, \"%s\", %d) == 0", spec.ProcessName, len(spec.ProcessName)))
+	}
+
+	// Add custom filter
+	if spec.Filter != "" {
+		// Convert common patterns to bpftrace syntax
+		customFilter := strings.ReplaceAll(spec.Filter, "arg", "arg")
+		filters = append(filters, customFilter)
+	}
+
+	return filters
+}
+
+// buildOutputFormat creates the output format string
+func (tm *BCCTraceManager) buildOutputFormat(spec TraceSpec) string {
+	if spec.Format != "" {
+		// Use custom format
+		return fmt.Sprintf("TRACE|%%d|%%d|%%d|%%s|%s|%s", spec.Target, spec.Format)
+	}
+
+	// Default format
+	return fmt.Sprintf("TRACE|%%d|%%d|%%d|%%s|%s|called", spec.Target)
+}
+
+// buildArgumentList creates the argument list for printf
+func (tm *BCCTraceManager) buildArgumentList(spec TraceSpec) []string {
+	// Always include timestamp, pid, tid, comm
+	args := []string{"nsecs", "pid", "tid", "comm"}
+
+	// Add custom arguments
+	for _, arg := range spec.Arguments {
+		switch arg {
+		case "arg1", "arg2", "arg3", "arg4", "arg5", "arg6":
+			args = append(args, fmt.Sprintf("arg%s", strings.TrimPrefix(arg, "arg")))
+		case "retval":
+			args = append(args, "retval")
+		case "cpu":
+			args = append(args, "cpu")
+		default:
+			// Custom expression
+			args = append(args, arg)
+		}
+	}
+
+	return args
+}
+
+// monitorTrace monitors a running trace and collects events
+func (tm *BCCTraceManager) monitorTrace(traceID string, stdout io.ReadCloser) {
+	tm.tracesLock.Lock()
+	trace, exists := tm.traces[traceID]
+	if !exists {
+		tm.tracesLock.Unlock()
+		return
+	}
+	tm.tracesLock.Unlock()
+
+	// Start reading output in a goroutine
+	go func() {
+		scanner := NewEventScanner(stdout)
+		for scanner.Scan() {
+			event := scanner.Event()
+			if event != nil {
+				tm.tracesLock.Lock()
+				if t, exists := tm.traces[traceID]; exists {
+					t.Events = append(t.Events, *event)
+				}
+				tm.tracesLock.Unlock()
+			}
+		}
+		stdout.Close()
+	}()
+
+	// Wait for the process to complete
+	err := trace.Process.Wait()
+
+	// Clean up
+	trace.Cancel()
+
+	tm.tracesLock.Lock()
+	if err != nil && err.Error() != "signal: killed" {
+		logging.Warning("Trace %s completed with error: %v", traceID, err)
+	} else {
+		logging.Debug("Trace %s completed successfully with %d events",
+			traceID, len(trace.Events))
+	}
+
+	// Signal that monitoring is complete
+	close(trace.Done)
+	tm.tracesLock.Unlock()
+}
+
+// GetTraceResult returns the results of a completed trace
+func (tm *BCCTraceManager) GetTraceResult(traceID string) (*TraceResult, error) {
+	tm.tracesLock.RLock()
+	trace, exists := tm.traces[traceID]
+	if !exists {
+		tm.tracesLock.RUnlock()
+		return nil, fmt.Errorf("trace %s not found", traceID)
+	}
+	tm.tracesLock.RUnlock()
+
+	// Wait for trace monitoring to complete
+	select {
+	case <-trace.Done:
+		// Trace monitoring completed
+	case <-time.After(5 * time.Second):
+		// Timeout waiting for completion
+		return nil, fmt.Errorf("timeout waiting for trace %s to complete", traceID)
+	}
+
+	// Now safely read the final results
+	tm.tracesLock.RLock()
+	defer tm.tracesLock.RUnlock()
+
+	result := &TraceResult{
+		TraceID:    traceID,
+		Spec:       trace.Spec,
+		Events:     make([]TraceEvent, len(trace.Events)),
+		EventCount: len(trace.Events),
+		StartTime:  trace.StartTime,
+		EndTime:    time.Now(),
+	}
+
+	copy(result.Events, trace.Events)
+
+	// Calculate statistics
+	result.Statistics = tm.calculateStatistics(result.Events, result.EndTime.Sub(result.StartTime))
+
+	// Generate summary
+	result.Summary = tm.generateSummary(result)
+
+	return result, nil
+}
+
+// calculateStatistics calculates statistics for the trace results
+func (tm *BCCTraceManager) calculateStatistics(events []TraceEvent, duration time.Duration) TraceStats {
+	stats := TraceStats{
+		TotalEvents:     len(events),
+		EventsByProcess: make(map[string]int),
+		EventsByUID:     make(map[int]int),
+	}
+
+	if duration > 0 {
+		stats.EventsPerSecond = float64(len(events)) / duration.Seconds()
+	}
+
+	// Calculate per-process and per-UID statistics
+	for _, event := range events {
+		stats.EventsByProcess[event.ProcessName]++
+		stats.EventsByUID[event.UID]++
+	}
+
+	// Calculate top processes
+	for processName, count := range stats.EventsByProcess {
+		percentage := float64(count) / float64(len(events)) * 100
+		stats.TopProcesses = append(stats.TopProcesses, ProcessStat{
+			ProcessName: processName,
+			EventCount:  count,
+			Percentage:  percentage,
+		})
+	}
+
+	return stats
+}
+
+// generateSummary generates a human-readable summary
+func (tm *BCCTraceManager) generateSummary(result *TraceResult) string {
+	duration := result.EndTime.Sub(result.StartTime)
+
+	summary := fmt.Sprintf("Traced %s for %v, captured %d events (%.2f events/sec)",
+		result.Spec.Target, duration, result.EventCount, result.Statistics.EventsPerSecond)
+
+	if len(result.Statistics.TopProcesses) > 0 {
+		summary += fmt.Sprintf(", top process: %s (%d events)",
+			result.Statistics.TopProcesses[0].ProcessName,
+			result.Statistics.TopProcesses[0].EventCount)
+	}
+
+	return summary
+}
+
+// StopTrace stops an active trace
+func (tm *BCCTraceManager) StopTrace(traceID string) error {
+	tm.tracesLock.Lock()
+	defer tm.tracesLock.Unlock()
+
+	trace, exists := tm.traces[traceID]
+	if !exists {
+		return fmt.Errorf("trace %s not found", traceID)
+	}
+
+	if trace.Process.ProcessState == nil {
+		// Process is still running, kill it
+		if err := trace.Process.Process.Kill(); err != nil {
+			return fmt.Errorf("failed to stop trace: %w", err)
+		}
+	}
+
+	trace.Cancel()
+	return nil
+}
+
+// ListActiveTraces returns a list of active trace IDs
+func (tm *BCCTraceManager) ListActiveTraces() []string {
+	tm.tracesLock.RLock()
+	defer tm.tracesLock.RUnlock()
+
+	var active []string
+	for id, trace := range tm.traces {
+		if trace.Process.ProcessState == nil {
+			active = append(active, id)
+		}
+	}
+
+	return active
+}
+
+// GetSummary returns a summary of the trace manager state
+func (tm *BCCTraceManager) GetSummary() map[string]interface{} {
+	tm.tracesLock.RLock()
+	defer tm.tracesLock.RUnlock()
+
+	activeCount := 0
+	completedCount := 0
+
+	for _, trace := range tm.traces {
+		if trace.Process.ProcessState == nil {
+			activeCount++
+		} else {
+			completedCount++
+		}
+	}
+
+	return map[string]interface{}{
+		"capabilities":     tm.capabilities,
+		"active_traces":    activeCount,
+		"completed_traces": completedCount,
+		"total_traces":     len(tm.traces),
+		"active_trace_ids": tm.ListActiveTraces(),
+	}
+}

ebpf_trace_specs.go

@@ -0,0 +1,396 @@
+package main
+
+import (
+	"encoding/json"
+	"fmt"
+	"strings"
+)
+
+// TestTraceSpecs provides test trace specifications for unit testing the BCC-style tracing
+// These are used to validate the tracing functionality without requiring remote API calls
+var TestTraceSpecs = map[string]TraceSpec{
+	// Basic system call tracing for testing
+	"test_sys_open": {
+		ProbeType: "p",
+		Target:    "__x64_sys_openat",
+		Format:    "opening file: %s",
+		Arguments: []string{"arg2@user"}, // filename
+		Duration:  5,                     // Short duration for testing
+	},
+
+	"test_sys_read": {
+		ProbeType: "p",
+		Target:    "__x64_sys_read",
+		Format:    "read %d bytes from fd %d",
+		Arguments: []string{"arg3", "arg1"}, // count, fd
+		Filter:    "arg3 > 100",             // Only reads >100 bytes for testing
+		Duration:  5,
+	},
+
+	"test_sys_write": {
+		ProbeType: "p",
+		Target:    "__x64_sys_write",
+		Format:    "write %d bytes to fd %d",
+		Arguments: []string{"arg3", "arg1"}, // count, fd
+		Duration:  5,
+	},
+
+	"test_process_creation": {
+		ProbeType: "p",
+		Target:    "__x64_sys_execve",
+		Format:    "exec: %s",
+		Arguments: []string{"arg1@user"}, // filename
+		Duration:  5,
+	},
+
+	// Test with different probe types
+	"test_kretprobe": {
+		ProbeType: "r",
+		Target:    "__x64_sys_openat",
+		Format:    "open returned: %d",
+		Arguments: []string{"retval"},
+		Duration:  5,
+	},
+
+	"test_with_filter": {
+		ProbeType: "p",
+		Target:    "__x64_sys_write",
+		Format:    "stdout write: %d bytes",
+		Arguments: []string{"arg3"},
+		Filter:    "arg1 == 1", // Only stdout writes
+		Duration:  5,
+	},
+}
+
+// GetTestSpec returns a pre-defined test trace specification
+func GetTestSpec(name string) (TraceSpec, bool) {
+	spec, exists := TestTraceSpecs[name]
+	return spec, exists
+}
+
+// ListTestSpecs returns all available test trace specifications
+func ListTestSpecs() map[string]string {
+	descriptions := map[string]string{
+		"test_sys_open":         "Test file open operations",
+		"test_sys_read":         "Test read operations (>100 bytes)",
+		"test_sys_write":        "Test write operations",
+		"test_process_creation": "Test process execution",
+		"test_kretprobe":        "Test kretprobe on file open",
+		"test_with_filter":      "Test filtered writes to stdout",
+	}
+
+	return descriptions
+}
+
+// TraceSpecBuilder helps build custom trace specifications
+type TraceSpecBuilder struct {
+	spec TraceSpec
+}
+
+// NewTraceSpecBuilder creates a new trace specification builder
+func NewTraceSpecBuilder() *TraceSpecBuilder {
+	return &TraceSpecBuilder{
+		spec: TraceSpec{
+			ProbeType: "p", // Default to kprobe
+			Duration:  30,  // Default 30 seconds
+		},
+	}
+}
+
+// Kprobe sets up a kernel probe
+func (b *TraceSpecBuilder) Kprobe(function string) *TraceSpecBuilder {
+	b.spec.ProbeType = "p"
+	b.spec.Target = function
+	return b
+}
+
+// Kretprobe sets up a kernel return probe
+func (b *TraceSpecBuilder) Kretprobe(function string) *TraceSpecBuilder {
+	b.spec.ProbeType = "r"
+	b.spec.Target = function
+	return b
+}
+
+// Tracepoint sets up a tracepoint
+func (b *TraceSpecBuilder) Tracepoint(category, name string) *TraceSpecBuilder {
+	b.spec.ProbeType = "t"
+	b.spec.Target = fmt.Sprintf("%s:%s", category, name)
+	return b
+}
+
+// Uprobe sets up a userspace probe
+func (b *TraceSpecBuilder) Uprobe(library, function string) *TraceSpecBuilder {
+	b.spec.ProbeType = "u"
+	b.spec.Library = library
+	b.spec.Target = function
+	return b
+}
+
+// Format sets the output format string
+func (b *TraceSpecBuilder) Format(format string, args ...string) *TraceSpecBuilder {
+	b.spec.Format = format
+	b.spec.Arguments = args
+	return b
+}
+
+// Filter adds a filter condition
+func (b *TraceSpecBuilder) Filter(condition string) *TraceSpecBuilder {
+	b.spec.Filter = condition
+	return b
+}
+
+// Duration sets the trace duration in seconds
+func (b *TraceSpecBuilder) Duration(seconds int) *TraceSpecBuilder {
+	b.spec.Duration = seconds
+	return b
+}
+
+// PID filters by process ID
+func (b *TraceSpecBuilder) PID(pid int) *TraceSpecBuilder {
+	b.spec.PID = pid
+	return b
+}
+
+// UID filters by user ID
+func (b *TraceSpecBuilder) UID(uid int) *TraceSpecBuilder {
+	b.spec.UID = uid
+	return b
+}
+
+// ProcessName filters by process name
+func (b *TraceSpecBuilder) ProcessName(name string) *TraceSpecBuilder {
+	b.spec.ProcessName = name
+	return b
+}
+
+// Build returns the constructed trace specification
+func (b *TraceSpecBuilder) Build() TraceSpec {
+	return b.spec
+}
+
+// TraceSpecParser parses trace specifications from various formats
+type TraceSpecParser struct{}
+
+// NewTraceSpecParser creates a new parser
+func NewTraceSpecParser() *TraceSpecParser {
+	return &TraceSpecParser{}
+}
+
+// ParseFromBCCStyle parses BCC trace.py style specifications
+// Examples:
+//
+//	"sys_open" -> trace sys_open syscall
+//	"p::do_sys_open" -> kprobe on do_sys_open
+//	"r::do_sys_open" -> kretprobe on do_sys_open
+//	"t:syscalls:sys_enter_open" -> tracepoint
+//	"sys_read (arg3 > 1024)" -> with filter
+//	"sys_read \"read %d bytes\", arg3" -> with format
+func (p *TraceSpecParser) ParseFromBCCStyle(spec string) (TraceSpec, error) {
+	result := TraceSpec{
+		ProbeType: "p",
+		Duration:  30,
+	}
+
+	// Split by quotes to separate format string
+	parts := strings.Split(spec, "\"")
+
+	var probeSpec string
+	if len(parts) >= 1 {
+		probeSpec = strings.TrimSpace(parts[0])
+	}
+
+	var formatPart string
+	if len(parts) >= 2 {
+		formatPart = parts[1]
+	}
+
+	var argsPart string
+	if len(parts) >= 3 {
+		argsPart = strings.TrimSpace(parts[2])
+		if strings.HasPrefix(argsPart, ",") {
+			argsPart = strings.TrimSpace(argsPart[1:])
+		}
+	}
+
+	// Parse probe specification
+	if err := p.parseProbeSpec(probeSpec, &result); err != nil {
+		return result, err
+	}
+
+	// Parse format string
+	if formatPart != "" {
+		result.Format = formatPart
+	}
+
+	// Parse arguments
+	if argsPart != "" {
+		result.Arguments = p.parseArguments(argsPart)
+	}
+
+	return result, nil
+}
+
+// parseProbeSpec parses the probe specification part
+func (p *TraceSpecParser) parseProbeSpec(spec string, result *TraceSpec) error {
+	// Handle filter conditions in parentheses
+	if idx := strings.Index(spec, "("); idx != -1 {
+		filterEnd := strings.LastIndex(spec, ")")
+		if filterEnd > idx {
+			result.Filter = strings.TrimSpace(spec[idx+1 : filterEnd])
+			spec = strings.TrimSpace(spec[:idx])
+		}
+	}
+
+	// Parse probe type and target
+	if strings.Contains(spec, ":") {
+		parts := strings.SplitN(spec, ":", 3)
+
+		if len(parts) >= 1 && parts[0] != "" {
+			switch parts[0] {
+			case "p":
+				result.ProbeType = "p"
+			case "r":
+				result.ProbeType = "r"
+			case "t":
+				result.ProbeType = "t"
+			case "u":
+				result.ProbeType = "u"
+			default:
+				return fmt.Errorf("unsupported probe type: %s", parts[0])
+			}
+		}
+
+		if len(parts) >= 2 {
+			result.Library = parts[1]
+		}
+
+		if len(parts) >= 3 {
+			result.Target = parts[2]
+		} else if len(parts) == 2 {
+			result.Target = parts[1]
+			result.Library = ""
+		}
+	} else {
+		// Simple function name
+		result.Target = spec
+
+		// Auto-detect syscall format
+		if strings.HasPrefix(spec, "sys_") && !strings.HasPrefix(spec, "__x64_sys_") {
+			result.Target = "__x64_sys_" + spec[4:]
+		}
+	}
+
+	return nil
+}
+
+// parseArguments parses the arguments part
+func (p *TraceSpecParser) parseArguments(args string) []string {
+	var result []string
+
+	// Split by comma and clean up
+	parts := strings.Split(args, ",")
+	for _, part := range parts {
+		arg := strings.TrimSpace(part)
+		if arg != "" {
+			result = append(result, arg)
+		}
+	}
+
+	return result
+}
+
+// ParseFromJSON parses trace specification from JSON
+func (p *TraceSpecParser) ParseFromJSON(jsonData []byte) (TraceSpec, error) {
+	var spec TraceSpec
+	err := json.Unmarshal(jsonData, &spec)
+	return spec, err
+}
+
+// GetCommonSpec returns a pre-defined test trace specification (renamed for backward compatibility)
+func GetCommonSpec(name string) (TraceSpec, bool) {
+	// Map old names to new test names for compatibility
+	testName := name
+	if strings.HasPrefix(name, "trace_") {
+		testName = strings.Replace(name, "trace_", "test_", 1)
+	}
+
+	spec, exists := TestTraceSpecs[testName]
+	return spec, exists
+}
+
+// ListCommonSpecs returns all available test trace specifications (renamed for backward compatibility)
+func ListCommonSpecs() map[string]string {
+	return ListTestSpecs()
+}
+
+// ValidateTraceSpec validates a trace specification
+func ValidateTraceSpec(spec TraceSpec) error {
+	if spec.Target == "" {
+		return fmt.Errorf("target function/syscall is required")
+	}
+
+	if spec.Duration <= 0 {
+		return fmt.Errorf("duration must be positive")
+	}
+
+	if spec.Duration > 600 { // 10 minutes max
+		return fmt.Errorf("duration too long (max 600 seconds)")
+	}
+
+	switch spec.ProbeType {
+	case "p", "r", "t", "u":
+		// Valid probe types
+	case "":
+		// Default to kprobe
+	default:
+		return fmt.Errorf("unsupported probe type: %s", spec.ProbeType)
+	}
+
+	if spec.ProbeType == "u" && spec.Library == "" {
+		return fmt.Errorf("library required for userspace probes")
+	}
+
+	if spec.ProbeType == "t" && !strings.Contains(spec.Target, ":") {
+		return fmt.Errorf("tracepoint requires format 'category:name'")
+	}
+
+	return nil
+}
+
+// SuggestSyscallTargets suggests syscall targets based on the issue description
+func SuggestSyscallTargets(issueDescription string) []string {
+	description := strings.ToLower(issueDescription)
+	var suggestions []string
+
+	// File I/O issues
+	if strings.Contains(description, "file") || strings.Contains(description, "disk") || strings.Contains(description, "io") {
+		suggestions = append(suggestions, "trace_sys_open", "trace_sys_read", "trace_sys_write", "trace_sys_unlink")
+	}
+
+	// Network issues
+	if strings.Contains(description, "network") || strings.Contains(description, "socket") || strings.Contains(description, "connection") {
+		suggestions = append(suggestions, "trace_sys_connect", "trace_sys_socket", "trace_sys_bind", "trace_sys_accept")
+	}
+
+	// Process issues
+	if strings.Contains(description, "process") || strings.Contains(description, "crash") || strings.Contains(description, "exec") {
+		suggestions = append(suggestions, "trace_sys_execve", "trace_sys_clone", "trace_sys_exit", "trace_sys_kill")
+	}
+
+	// Memory issues
+	if strings.Contains(description, "memory") || strings.Contains(description, "malloc") || strings.Contains(description, "leak") {
+		suggestions = append(suggestions, "trace_sys_mmap", "trace_sys_brk")
+	}
+
+	// Performance issues - trace common syscalls
+	if strings.Contains(description, "slow") || strings.Contains(description, "performance") || strings.Contains(description, "hang") {
+		suggestions = append(suggestions, "trace_sys_read", "trace_sys_write", "trace_sys_connect", "trace_sys_mmap")
+	}
+
+	// If no specific suggestions, provide general monitoring
+	if len(suggestions) == 0 {
+		suggestions = append(suggestions, "trace_sys_execve", "trace_sys_open", "trace_sys_connect")
+	}
+
+	return suggestions
+}

ebpf_trace_test.go

@@ -0,0 +1,878 @@
+package main
+
+import (
+	"encoding/json"
+	"fmt"
+	"os"
+	"strings"
+	"testing"
+	"time"
+)
+
+// TestBCCTracing demonstrates and tests the new BCC-style tracing functionality
+// This test documents the expected behavior and response format of the agent
+func TestBCCTracing(t *testing.T) {
+	fmt.Println("=== BCC-Style eBPF Tracing Unit Tests ===")
+	fmt.Println()
+
+	// Test 1: List available test specifications
+	t.Run("ListTestSpecs", func(t *testing.T) {
+		specs := ListTestSpecs()
+		fmt.Printf("📋 Available Test Specifications:\n")
+		for name, description := range specs {
+			fmt.Printf("   - %s: %s\n", name, description)
+		}
+		fmt.Println()
+
+		if len(specs) == 0 {
+			t.Error("No test specifications available")
+		}
+	})
+
+	// Test 2: Parse BCC-style specifications
+	t.Run("ParseBCCStyle", func(t *testing.T) {
+		parser := NewTraceSpecParser()
+
+		testCases := []struct {
+			input    string
+			expected string
+		}{
+			{
+				input:    "sys_open",
+				expected: "__x64_sys_open",
+			},
+			{
+				input:    "p::do_sys_open",
+				expected: "do_sys_open",
+			},
+			{
+				input:    "r::sys_read",
+				expected: "sys_read",
+			},
+			{
+				input:    "sys_write (arg1 == 1)",
+				expected: "__x64_sys_write",
+			},
+		}
+
+		fmt.Printf("🔍 Testing BCC-style parsing:\n")
+		for _, tc := range testCases {
+			spec, err := parser.ParseFromBCCStyle(tc.input)
+			if err != nil {
+				t.Errorf("Failed to parse '%s': %v", tc.input, err)
+				continue
+			}
+
+			fmt.Printf("   Input: '%s' -> Target: '%s', Type: '%s'\n",
+				tc.input, spec.Target, spec.ProbeType)
+
+			if spec.Target != tc.expected {
+				t.Errorf("Expected target '%s', got '%s'", tc.expected, spec.Target)
+			}
+		}
+		fmt.Println()
+	})
+
+	// Test 3: Validate trace specifications
+	t.Run("ValidateSpecs", func(t *testing.T) {
+		fmt.Printf("✅ Testing trace specification validation:\n")
+
+		// Valid spec
+		validSpec := TraceSpec{
+			ProbeType: "p",
+			Target:    "__x64_sys_openat",
+			Format:    "opening file",
+			Duration:  5,
+		}
+
+		if err := ValidateTraceSpec(validSpec); err != nil {
+			t.Errorf("Valid spec failed validation: %v", err)
+		} else {
+			fmt.Printf("   ✓ Valid specification passed\n")
+		}
+
+		// Invalid spec - no target
+		invalidSpec := TraceSpec{
+			ProbeType: "p",
+			Duration:  5,
+		}
+
+		if err := ValidateTraceSpec(invalidSpec); err == nil {
+			t.Error("Invalid spec (no target) should have failed validation")
+		} else {
+			fmt.Printf("   ✓ Invalid specification correctly rejected: %s\n", err.Error())
+		}
+
+		fmt.Println()
+	})
+
+	// Test 4: Simulate agent response format
+	t.Run("SimulateAgentResponse", func(t *testing.T) {
+		fmt.Printf("🤖 Simulating agent response for BCC-style tracing:\n")
+
+		// Get a test specification
+		testSpec, exists := GetTestSpec("test_sys_open")
+		if !exists {
+			t.Fatal("test_sys_open specification not found")
+		}
+
+		// Simulate what the agent would return
+		mockResponse := simulateTraceExecution(testSpec)
+
+		// Print the response format
+		responseJSON, _ := json.MarshalIndent(mockResponse, "", "  ")
+		fmt.Printf("   Expected Response Format:\n%s\n", string(responseJSON))
+
+		// Validate response structure
+		if mockResponse["success"] != true {
+			t.Error("Expected successful trace execution")
+		}
+
+		if mockResponse["type"] != "bcc_trace" {
+			t.Error("Expected type to be 'bcc_trace'")
+		}
+
+		events, hasEvents := mockResponse["events"].([]TraceEvent)
+		if !hasEvents || len(events) == 0 {
+			t.Error("Expected trace events in response")
+		}
+
+		fmt.Println()
+	})
+
+	// Test 5: Test different probe types
+	t.Run("TestProbeTypes", func(t *testing.T) {
+		fmt.Printf("🔬 Testing different probe types:\n")
+
+		probeTests := []struct {
+			specName string
+			expected string
+		}{
+			{"test_sys_open", "kprobe"},
+			{"test_kretprobe", "kretprobe"},
+			{"test_with_filter", "kprobe with filter"},
+		}
+
+		for _, test := range probeTests {
+			spec, exists := GetTestSpec(test.specName)
+			if !exists {
+				t.Errorf("Test spec '%s' not found", test.specName)
+				continue
+			}
+
+			response := simulateTraceExecution(spec)
+			fmt.Printf("   %s -> %s: %d events captured\n",
+				test.specName, test.expected, response["event_count"])
+		}
+		fmt.Println()
+	})
+
+	// Test 6: Test trace spec builder
+	t.Run("TestTraceSpecBuilder", func(t *testing.T) {
+		fmt.Printf("🏗️  Testing trace specification builder:\n")
+
+		// Build a custom trace spec
+		spec := NewTraceSpecBuilder().
+			Kprobe("__x64_sys_write").
+			Format("write syscall: %d bytes", "arg3").
+			Filter("arg1 == 1").
+			Duration(3).
+			Build()
+
+		fmt.Printf("   Built spec: Target=%s, Format=%s, Filter=%s\n",
+			spec.Target, spec.Format, spec.Filter)
+
+		if spec.Target != "__x64_sys_write" {
+			t.Error("Builder failed to set target correctly")
+		}
+
+		if spec.ProbeType != "p" {
+			t.Error("Builder failed to set probe type correctly")
+		}
+
+		fmt.Println()
+	})
+}
+
+// simulateTraceExecution simulates what the agent would return for a trace execution
+// This documents the expected response format from the agent
+func simulateTraceExecution(spec TraceSpec) map[string]interface{} {
+	// Simulate some trace events
+	events := []TraceEvent{
+		{
+			Timestamp:   time.Now().Unix(),
+			PID:         1234,
+			TID:         1234,
+			ProcessName: "test_process",
+			Function:    spec.Target,
+			Message:     fmt.Sprintf(spec.Format, "test_file.txt"),
+			RawArgs: map[string]string{
+				"arg1": "5",
+				"arg2": "test_file.txt",
+				"arg3": "1024",
+			},
+		},
+		{
+			Timestamp:   time.Now().Unix(),
+			PID:         5678,
+			TID:         5678,
+			ProcessName: "another_process",
+			Function:    spec.Target,
+			Message:     fmt.Sprintf(spec.Format, "data.log"),
+			RawArgs: map[string]string{
+				"arg1": "3",
+				"arg2": "data.log",
+				"arg3": "512",
+			},
+		},
+	}
+
+	// Simulate trace statistics
+	stats := TraceStats{
+		TotalEvents:     len(events),
+		EventsByProcess: map[string]int{"test_process": 1, "another_process": 1},
+		EventsByUID:     map[int]int{1000: 2},
+		EventsPerSecond: float64(len(events)) / float64(spec.Duration),
+		TopProcesses: []ProcessStat{
+			{ProcessName: "test_process", EventCount: 1, Percentage: 50.0},
+			{ProcessName: "another_process", EventCount: 1, Percentage: 50.0},
+		},
+	}
+
+	// Return the expected agent response format
+	return map[string]interface{}{
+		"name":        spec.Target,
+		"type":        "bcc_trace",
+		"target":      spec.Target,
+		"duration":    spec.Duration,
+		"description": fmt.Sprintf("Traced %s for %d seconds", spec.Target, spec.Duration),
+		"status":      "completed",
+		"success":     true,
+		"event_count": len(events),
+		"events":      events,
+		"statistics":  stats,
+		"data_points": len(events),
+		"probe_type":  spec.ProbeType,
+		"format":      spec.Format,
+		"filter":      spec.Filter,
+	}
+}
+
+// TestTraceManagerCapabilities tests the trace manager capabilities
+func TestTraceManagerCapabilities(t *testing.T) {
+	fmt.Println("=== BCC Trace Manager Capabilities Test ===")
+	fmt.Println()
+
+	manager := NewBCCTraceManager()
+	caps := manager.GetCapabilities()
+
+	fmt.Printf("🔧 Trace Manager Capabilities:\n")
+	for capability, available := range caps {
+		status := "❌ Not Available"
+		if available {
+			status = "✅ Available"
+		}
+		fmt.Printf("   %s: %s\n", capability, status)
+	}
+	fmt.Println()
+
+	// Check essential capabilities
+	if !caps["kernel_ebpf"] {
+		fmt.Printf("⚠️  Warning: Kernel eBPF support not detected\n")
+	}
+
+	if !caps["bpftrace"] {
+		fmt.Printf("⚠️  Warning: bpftrace not available (install with: apt install bpftrace)\n")
+	}
+
+	if !caps["root_access"] {
+		fmt.Printf("⚠️  Warning: Root access required for eBPF tracing\n")
+	}
+}
+
+// BenchmarkTraceSpecParsing benchmarks the trace specification parsing
+func BenchmarkTraceSpecParsing(b *testing.B) {
+	parser := NewTraceSpecParser()
+	testInput := "sys_open \"opening %s\", arg2@user"
+
+	b.ResetTimer()
+	for i := 0; i < b.N; i++ {
+		_, err := parser.ParseFromBCCStyle(testInput)
+		if err != nil {
+			b.Fatal(err)
+		}
+	}
+}
+
+// TestSyscallSuggestions tests the syscall suggestion functionality
+func TestSyscallSuggestions(t *testing.T) {
+	fmt.Println("=== Syscall Suggestion Test ===")
+	fmt.Println()
+
+	testCases := []struct {
+		issue       string
+		expected    int // minimum expected suggestions
+		description string
+	}{
+		{
+			issue:       "file not found error",
+			expected:    1,
+			description: "File I/O issue should suggest file-related syscalls",
+		},
+		{
+			issue:       "network connection timeout",
+			expected:    1,
+			description: "Network issue should suggest network syscalls",
+		},
+		{
+			issue:       "process crashes randomly",
+			expected:    1,
+			description: "Process issue should suggest process-related syscalls",
+		},
+		{
+			issue:       "memory leak detected",
+			expected:    1,
+			description: "Memory issue should suggest memory syscalls",
+		},
+		{
+			issue:       "application is slow",
+			expected:    1,
+			description: "Performance issue should suggest monitoring syscalls",
+		},
+	}
+
+	fmt.Printf("💡 Testing syscall suggestions:\n")
+	for _, tc := range testCases {
+		suggestions := SuggestSyscallTargets(tc.issue)
+		fmt.Printf("   Issue: '%s' -> %d suggestions: %v\n",
+			tc.issue, len(suggestions), suggestions)
+
+		if len(suggestions) < tc.expected {
+			t.Errorf("Expected at least %d suggestions for '%s', got %d",
+				tc.expected, tc.issue, len(suggestions))
+		}
+	}
+	fmt.Println()
+}
+
+// TestMain runs the tests and provides a summary
+func TestMain(m *testing.M) {
+	fmt.Println("🚀 Starting BCC-Style eBPF Tracing Tests")
+	fmt.Println("========================================")
+	fmt.Println()
+
+	// Run capability check first
+	manager := NewBCCTraceManager()
+	caps := manager.GetCapabilities()
+
+	if !caps["kernel_ebpf"] {
+		fmt.Println("⚠️  Kernel eBPF support not detected - some tests may be limited")
+	}
+	if !caps["bpftrace"] {
+		fmt.Println("⚠️  bpftrace not available - install with: sudo apt install bpftrace")
+	}
+	if !caps["root_access"] {
+		fmt.Println("⚠️  Root access required for actual eBPF tracing")
+	}
+
+	fmt.Println()
+
+	// Run the tests
+	code := m.Run()
+
+	fmt.Println()
+	fmt.Println("========================================")
+	if code == 0 {
+		fmt.Println("✅ All BCC-Style eBPF Tracing Tests Passed!")
+	} else {
+		fmt.Println("❌ Some tests failed")
+	}
+
+	os.Exit(code)
+}
+
+// TestBCCTraceManagerRootTest tests the actual BCC trace manager with root privileges
+// This test requires root access and will only run meaningful tests when root
+func TestBCCTraceManagerRootTest(t *testing.T) {
+	fmt.Println("=== BCC Trace Manager Root Test ===")
+
+	// Check if running as root
+	if os.Geteuid() != 0 {
+		t.Skip("⚠️  Skipping root test - not running as root (use: sudo go test -run TestBCCTraceManagerRootTest)")
+		return
+	}
+
+	fmt.Println("✅ Running as root - can test actual eBPF functionality")
+
+	// Test 1: Create BCC trace manager and check capabilities
+	manager := NewBCCTraceManager()
+	caps := manager.GetCapabilities()
+
+	fmt.Printf("🔍 BCC Trace Manager Capabilities:\n")
+	for cap, available := range caps {
+		status := "❌"
+		if available {
+			status = "✅"
+		}
+		fmt.Printf("   %s %s: %v\n", status, cap, available)
+	}
+
+	// Require essential capabilities
+	if !caps["bpftrace"] {
+		t.Fatal("❌ bpftrace not available - install bpftrace package")
+	}
+
+	if !caps["root_access"] {
+		t.Fatal("❌ Root access not detected")
+	}
+
+	// Test 2: Create and execute a simple trace
+	fmt.Println("\n🔬 Testing actual eBPF trace execution...")
+
+	spec := TraceSpec{
+		ProbeType: "t", // tracepoint
+		Target:    "syscalls:sys_enter_openat",
+		Format:    "file access",
+		Arguments: []string{}, // Remove invalid arg2@user for tracepoints
+		Duration:  3,          // 3 seconds
+	}
+
+	fmt.Printf("📝 Starting trace: %s for %d seconds\n", spec.Target, spec.Duration)
+
+	traceID, err := manager.StartTrace(spec)
+	if err != nil {
+		t.Fatalf("❌ Failed to start trace: %v", err)
+	}
+
+	fmt.Printf("🚀 Trace started with ID: %s\n", traceID)
+
+	// Generate some file access to capture
+	go func() {
+		time.Sleep(1 * time.Second)
+		// Create some file operations to trace
+		for i := 0; i < 3; i++ {
+			testFile := fmt.Sprintf("/tmp/bcc_test_%d.txt", i)
+
+			// This will trigger sys_openat syscalls
+			if file, err := os.Create(testFile); err == nil {
+				file.WriteString("BCC trace test")
+				file.Close()
+				os.Remove(testFile)
+			}
+			time.Sleep(500 * time.Millisecond)
+		}
+	}()
+
+	// Wait for trace to complete
+	time.Sleep(time.Duration(spec.Duration+1) * time.Second)
+
+	// Get results
+	result, err := manager.GetTraceResult(traceID)
+	if err != nil {
+		// Try to stop the trace if it's still running
+		manager.StopTrace(traceID)
+		t.Fatalf("❌ Failed to get trace results: %v", err)
+	}
+
+	fmt.Printf("\n📊 Trace Results Summary:\n")
+	fmt.Printf("   • Trace ID: %s\n", result.TraceID)
+	fmt.Printf("   • Target: %s\n", result.Spec.Target)
+	fmt.Printf("   • Duration: %v\n", result.EndTime.Sub(result.StartTime))
+	fmt.Printf("   • Events captured: %d\n", result.EventCount)
+	fmt.Printf("   • Events per second: %.2f\n", result.Statistics.EventsPerSecond)
+	fmt.Printf("   • Summary: %s\n", result.Summary)
+
+	if len(result.Events) > 0 {
+		fmt.Printf("\n📝 Sample Events (first 3):\n")
+		for i, event := range result.Events {
+			if i >= 3 {
+				break
+			}
+			fmt.Printf("   %d. PID:%d TID:%d Process:%s Message:%s\n",
+				i+1, event.PID, event.TID, event.ProcessName, event.Message)
+		}
+
+		if len(result.Events) > 3 {
+			fmt.Printf("   ... and %d more events\n", len(result.Events)-3)
+		}
+	}
+
+	// Test 3: Validate the trace produced real data
+	if result.EventCount == 0 {
+		fmt.Println("⚠️  Warning: No events captured - this might be normal for a quiet system")
+	} else {
+		fmt.Printf("✅ Successfully captured %d real eBPF events!\n", result.EventCount)
+	}
+
+	fmt.Println("\n🧪 Testing comprehensive system tracing (Network, Disk, CPU, Memory, Userspace)...")
+
+	testSpecs := []TraceSpec{
+		// === SYSCALL TRACING ===
+		{
+			ProbeType: "p", // kprobe
+			Target:    "__x64_sys_write",
+			Format:    "write: fd=%d count=%d",
+			Arguments: []string{"arg1", "arg3"},
+			Duration:  2,
+		},
+		{
+			ProbeType: "p", // kprobe
+			Target:    "__x64_sys_read",
+			Format:    "read: fd=%d count=%d",
+			Arguments: []string{"arg1", "arg3"},
+			Duration:  2,
+		},
+		{
+			ProbeType: "p", // kprobe
+			Target:    "__x64_sys_connect",
+			Format:    "network connect: fd=%d",
+			Arguments: []string{"arg1"},
+			Duration:  2,
+		},
+		{
+			ProbeType: "p", // kprobe
+			Target:    "__x64_sys_accept",
+			Format:    "network accept: fd=%d",
+			Arguments: []string{"arg1"},
+			Duration:  2,
+		},
+		// === BLOCK I/O TRACING ===
+		{
+			ProbeType: "t", // tracepoint
+			Target:    "block:block_io_start",
+			Format:    "block I/O start",
+			Arguments: []string{},
+			Duration:  2,
+		},
+		{
+			ProbeType: "t", // tracepoint
+			Target:    "block:block_io_done",
+			Format:    "block I/O complete",
+			Arguments: []string{},
+			Duration:  2,
+		},
+		// === CPU SCHEDULER TRACING ===
+		{
+			ProbeType: "t", // tracepoint
+			Target:    "sched:sched_migrate_task",
+			Format:    "task migration",
+			Arguments: []string{},
+			Duration:  2,
+		},
+		{
+			ProbeType: "t", // tracepoint
+			Target:    "sched:sched_pi_setprio",
+			Format:    "priority change",
+			Arguments: []string{},
+			Duration:  2,
+		},
+		// === MEMORY MANAGEMENT ===
+		{
+			ProbeType: "t", // tracepoint
+			Target:    "syscalls:sys_enter_brk",
+			Format:    "memory allocation: brk",
+			Arguments: []string{},
+			Duration:  2,
+		},
+		// === KERNEL MEMORY TRACING ===
+		{
+			ProbeType: "t", // tracepoint
+			Target:    "kmem:kfree",
+			Format:    "kernel memory free",
+			Arguments: []string{},
+			Duration:  2,
+		},
+	}
+
+	for i, testSpec := range testSpecs {
+		category := "unknown"
+		if strings.Contains(testSpec.Target, "sys_write") || strings.Contains(testSpec.Target, "sys_read") {
+			category = "filesystem"
+		} else if strings.Contains(testSpec.Target, "sys_connect") || strings.Contains(testSpec.Target, "sys_accept") {
+			category = "network"
+		} else if strings.Contains(testSpec.Target, "block:") {
+			category = "disk I/O"
+		} else if strings.Contains(testSpec.Target, "sched:") {
+			category = "CPU/scheduler"
+		} else if strings.Contains(testSpec.Target, "sys_brk") || strings.Contains(testSpec.Target, "kmem:") {
+			category = "memory"
+		}
+
+		fmt.Printf("\n   🔍 Test %d: [%s] Tracing %s for %d seconds\n", i+1, category, testSpec.Target, testSpec.Duration)
+
+		testTraceID, err := manager.StartTrace(testSpec)
+		if err != nil {
+			fmt.Printf("   ❌ Failed to start: %v\n", err)
+			continue
+		}
+
+		// Generate activity specific to this trace type
+		go func(target, probeType string) {
+			time.Sleep(500 * time.Millisecond)
+			switch {
+			case strings.Contains(target, "sys_write") || strings.Contains(target, "sys_read"):
+				// Generate file I/O
+				for j := 0; j < 3; j++ {
+					testFile := fmt.Sprintf("/tmp/io_test_%d.txt", j)
+					if file, err := os.Create(testFile); err == nil {
+						file.WriteString("BCC tracing test data for I/O operations")
+						file.Sync()
+						file.Close()
+
+						// Read the file back
+						if readFile, err := os.Open(testFile); err == nil {
+							buffer := make([]byte, 1024)
+							readFile.Read(buffer)
+							readFile.Close()
+						}
+						os.Remove(testFile)
+					}
+					time.Sleep(200 * time.Millisecond)
+				}
+			case strings.Contains(target, "block:"):
+				// Generate disk I/O to trigger block layer events
+				for j := 0; j < 3; j++ {
+					testFile := fmt.Sprintf("/tmp/block_test_%d.txt", j)
+					if file, err := os.Create(testFile); err == nil {
+						// Write substantial data to trigger block I/O
+						data := make([]byte, 1024*4) // 4KB
+						for k := range data {
+							data[k] = byte(k % 256)
+						}
+						file.Write(data)
+						file.Sync() // Force write to disk
+						file.Close()
+					}
+					os.Remove(testFile)
+					time.Sleep(300 * time.Millisecond)
+				}
+			case strings.Contains(target, "sched:"):
+				// Generate CPU activity to trigger scheduler events
+				go func() {
+					for j := 0; j < 100; j++ {
+						// Create short-lived goroutines to trigger scheduler activity
+						go func() {
+							time.Sleep(time.Millisecond * 1)
+						}()
+						time.Sleep(time.Millisecond * 10)
+					}
+				}()
+			case strings.Contains(target, "sys_brk") || strings.Contains(target, "kmem:"):
+				// Generate memory allocation activity
+				for j := 0; j < 5; j++ {
+					// Allocate and free memory to trigger memory management
+					data := make([]byte, 1024*1024) // 1MB
+					for k := range data {
+						data[k] = byte(k % 256)
+					}
+					data = nil // Allow GC
+					time.Sleep(200 * time.Millisecond)
+				}
+			case strings.Contains(target, "sys_connect") || strings.Contains(target, "sys_accept"):
+				// Network operations (these may not generate events in test environment)
+				fmt.Printf("      Note: Network syscalls may not trigger events without actual network activity\n")
+			default:
+				// Generic activity
+				for j := 0; j < 3; j++ {
+					testFile := fmt.Sprintf("/tmp/generic_test_%d.txt", j)
+					if file, err := os.Create(testFile); err == nil {
+						file.WriteString("Generic test activity")
+						file.Close()
+					}
+					os.Remove(testFile)
+					time.Sleep(300 * time.Millisecond)
+				}
+			}
+		}(testSpec.Target, testSpec.ProbeType)
+
+		// Wait for trace completion
+		time.Sleep(time.Duration(testSpec.Duration+1) * time.Second)
+
+		testResult, err := manager.GetTraceResult(testTraceID)
+		if err != nil {
+			manager.StopTrace(testTraceID)
+			fmt.Printf("   ⚠️  Result error: %v\n", err)
+			continue
+		}
+
+		fmt.Printf("   📊 Results for %s:\n", testSpec.Target)
+		fmt.Printf("      • Total events: %d\n", testResult.EventCount)
+		fmt.Printf("      • Events/sec: %.2f\n", testResult.Statistics.EventsPerSecond)
+		fmt.Printf("      • Duration: %v\n", testResult.EndTime.Sub(testResult.StartTime))
+
+		// Show process breakdown
+		if len(testResult.Statistics.TopProcesses) > 0 {
+			fmt.Printf("      • Top processes:\n")
+			for j, proc := range testResult.Statistics.TopProcesses {
+				if j >= 3 { // Show top 3
+					break
+				}
+				fmt.Printf("        - %s: %d events (%.1f%%)\n",
+					proc.ProcessName, proc.EventCount, proc.Percentage)
+			}
+		}
+
+		// Show sample events with PIDs, counts, etc.
+		if len(testResult.Events) > 0 {
+			fmt.Printf("      • Sample events:\n")
+			for j, event := range testResult.Events {
+				if j >= 5 { // Show first 5 events
+					break
+				}
+				fmt.Printf("        [%d] PID:%d TID:%d Process:%s Message:%s\n",
+					j+1, event.PID, event.TID, event.ProcessName, event.Message)
+			}
+			if len(testResult.Events) > 5 {
+				fmt.Printf("        ... and %d more events\n", len(testResult.Events)-5)
+			}
+		}
+
+		if testResult.EventCount > 0 {
+			fmt.Printf("   ✅ Success: Captured %d real syscall events!\n", testResult.EventCount)
+		} else {
+			fmt.Printf("   ⚠️  No events captured (may be normal for this syscall)\n")
+		}
+	}
+
+	fmt.Println("\n🎉 BCC Trace Manager Root Test Complete!")
+	fmt.Println("✅ Real eBPF tracing is working and ready for production use!")
+}
+
+// TestAgentEBPFIntegration tests the agent's integration with BCC-style eBPF tracing
+// This demonstrates the complete flow from agent to eBPF results
+func TestAgentEBPFIntegration(t *testing.T) {
+	if os.Geteuid() != 0 {
+		t.Skip("⚠️  Skipping agent integration test - requires root access")
+		return
+	}
+
+	fmt.Println("\n=== Agent eBPF Integration Test ===")
+	fmt.Println("This test demonstrates the complete agent flow with BCC-style tracing")
+
+	// Create agent with eBPF manager
+	agent := &LinuxDiagnosticAgent{}
+	agent.ebpfManager = NewBCCTraceManager()
+	agent.config = DefaultAgentConfig() // Add config for concurrent execution
+
+	// Test multiple syscalls that would be sent by remote API
+	testEBPFRequests := []EBPFRequest{
+		{
+			Name:        "file_operations",
+			Type:        "syscall",
+			Target:      "sys_openat", // Will be converted to __x64_sys_openat
+			Duration:    3,
+			Description: "trace file open operations",
+			Filters:     map[string]string{},
+		},
+		{
+			Name:        "network_operations",
+			Type:        "syscall",
+			Target:      "__x64_sys_connect",
+			Duration:    2,
+			Description: "trace network connections",
+			Filters:     map[string]string{},
+		},
+		{
+			Name:        "io_operations",
+			Type:        "syscall",
+			Target:      "sys_write",
+			Duration:    2,
+			Description: "trace write operations",
+			Filters:     map[string]string{},
+		},
+	}
+
+	fmt.Printf("🚀 Testing agent with %d eBPF programs...\n\n", len(testEBPFRequests))
+
+	// Execute eBPF programs through agent (simulating API call)
+	traceSpecs := agent.convertEBPFProgramsToTraceSpecs(testEBPFRequests)
+	results := agent.executeBCCTracesConcurrently(traceSpecs)
+
+	fmt.Printf("📊 Agent eBPF Execution Results:\n")
+	fmt.Printf("=" + strings.Repeat("=", 50) + "\n\n")
+
+	for i, result := range results {
+		fmt.Printf("🔍 Program %d: %s\n", i+1, result["name"])
+		fmt.Printf("   Target: %s\n", result["target"])
+		fmt.Printf("   Type: %s\n", result["type"])
+		fmt.Printf("   Status: %s\n", result["status"])
+		fmt.Printf("   Success: %v\n", result["success"])
+
+		if result["success"].(bool) {
+			if eventCount, ok := result["event_count"].(int); ok {
+				fmt.Printf("   Events captured: %d\n", eventCount)
+			}
+			if dataPoints, ok := result["data_points"].(int); ok {
+				fmt.Printf("   Data points: %d\n", dataPoints)
+			}
+			if summary, ok := result["summary"].(string); ok {
+				fmt.Printf("   Summary: %s\n", summary)
+			}
+
+			// Show events if available
+			if events, ok := result["events"].([]TraceEvent); ok && len(events) > 0 {
+				fmt.Printf("   Sample events:\n")
+				for j, event := range events {
+					if j >= 3 { // Show first 3
+						break
+					}
+					fmt.Printf("     [%d] PID:%d Process:%s Message:%s\n",
+						j+1, event.PID, event.ProcessName, event.Message)
+				}
+				if len(events) > 3 {
+					fmt.Printf("     ... and %d more events\n", len(events)-3)
+				}
+			}
+
+			// Show statistics if available
+			if stats, ok := result["statistics"].(TraceStats); ok {
+				fmt.Printf("   Statistics:\n")
+				fmt.Printf("     - Events/sec: %.2f\n", stats.EventsPerSecond)
+				fmt.Printf("     - Total processes: %d\n", len(stats.EventsByProcess))
+				if len(stats.TopProcesses) > 0 {
+					fmt.Printf("     - Top process: %s (%d events)\n",
+						stats.TopProcesses[0].ProcessName, stats.TopProcesses[0].EventCount)
+				}
+			}
+		} else {
+			if errMsg, ok := result["error"].(string); ok {
+				fmt.Printf("   Error: %s\n", errMsg)
+			}
+		}
+		fmt.Println()
+	}
+
+	// Validate expected agent response format
+	t.Run("ValidateAgentResponseFormat", func(t *testing.T) {
+		for i, result := range results {
+			// Check required fields
+			requiredFields := []string{"name", "type", "target", "duration", "description", "status", "success"}
+			for _, field := range requiredFields {
+				if _, exists := result[field]; !exists {
+					t.Errorf("Result %d missing required field: %s", i, field)
+				}
+			}
+
+			// If successful, check for data fields
+			if success, ok := result["success"].(bool); ok && success {
+				// Should have either event_count or data_points
+				hasEventCount := false
+				hasDataPoints := false
+
+				if _, ok := result["event_count"]; ok {
+					hasEventCount = true
+				}
+				if _, ok := result["data_points"]; ok {
+					hasDataPoints = true
+				}
+
+				if !hasEventCount && !hasDataPoints {
+					t.Errorf("Successful result %d should have event_count or data_points", i)
+				}
+			}
+		}
+	})
+
+	fmt.Println("✅ Agent eBPF Integration Test Complete!")
+	fmt.Println("📈 The agent correctly processes eBPF requests and returns detailed syscall data!")
+}

go.mod

@@ -5,7 +5,7 @@ go 1.23.0
 toolchain go1.24.2
 
 require (
-	github.com/cilium/ebpf v0.19.0
+	github.com/gorilla/websocket v1.5.3
 	github.com/joho/godotenv v1.5.1
 	github.com/sashabaranov/go-openai v1.32.0
 	github.com/shirou/gopsutil/v3 v3.24.5
@@ -13,7 +13,6 @@ require (
 
 require (
 	github.com/go-ole/go-ole v1.2.6 // indirect
-	github.com/gorilla/websocket v1.5.3 // indirect
 	github.com/lufia/plan9stats v0.0.0-20211012122336-39d0f177ccd0 // indirect
 	github.com/power-devops/perfstat v0.0.0-20210106213030-5aafc221ea8c // indirect
 	github.com/shoenig/go-m1cpu v0.1.6 // indirect

go.sum

@@ -1,11 +1,7 @@
-github.com/cilium/ebpf v0.19.0 h1:Ro/rE64RmFBeA9FGjcTc+KmCeY6jXmryu6FfnzPRIao=
-github.com/cilium/ebpf v0.19.0/go.mod h1:fLCgMo3l8tZmAdM3B2XqdFzXBpwkcSTroaVqN08OWVY=
 github.com/davecgh/go-spew v1.1.1 h1:vj9j/u1bqnvCEfJOwUhtlOARqs3+rkHYY13jYWTU97c=
 github.com/davecgh/go-spew v1.1.1/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
 github.com/go-ole/go-ole v1.2.6 h1:/Fpf6oFPoeFik9ty7siob0G6Ke8QvQEuVcuChpwXzpY=
 github.com/go-ole/go-ole v1.2.6/go.mod h1:pprOEPIfldk/42T2oK7lQ4v4JSDwmV0As9GaiUsvbm0=
-github.com/go-quicktest/qt v1.101.1-0.20240301121107-c6c8733fa1e6 h1:teYtXy9B7y5lHTp8V9KPxpYRAVA7dozigQcMiBust1s=
-github.com/go-quicktest/qt v1.101.1-0.20240301121107-c6c8733fa1e6/go.mod h1:p4lGIVX+8Wa6ZPNDvqcxq36XpUDLh42FLetFU7odllI=
 github.com/google/go-cmp v0.5.6/go.mod h1:v8dTdLbMG2kIc/vJvl+f65V22dbkXbowE6jgT/gNBxE=
 github.com/google/go-cmp v0.6.0 h1:ofyhxvXcZhMsU5ulbFiLKl/XBFqE1GSq7atu8tAmTRI=
 github.com/google/go-cmp v0.6.0/go.mod h1:17dUlkBOakJ0+DkrSSNjCkIjxS6bF9zb3elmeNGIjoY=
@@ -13,26 +9,12 @@ github.com/gorilla/websocket v1.5.3 h1:saDtZ6Pbx/0u+bgYQ3q96pZgCzfhKXGPqt7kZ72aN
 github.com/gorilla/websocket v1.5.3/go.mod h1:YR8l580nyteQvAITg2hZ9XVh4b55+EU/adAjf1fMHhE=
 github.com/joho/godotenv v1.5.1 h1:7eLL/+HRGLY0ldzfGMeQkb7vMd0as4CfYvUVzLqw0N0=
 github.com/joho/godotenv v1.5.1/go.mod h1:f4LDr5Voq0i2e/R5DDNOoa2zzDfwtkZa6DnEwAbqwq4=
-github.com/josharian/native v1.1.0 h1:uuaP0hAbW7Y4l0ZRQ6C9zfb7Mg1mbFKry/xzDAfmtLA=
-github.com/josharian/native v1.1.0/go.mod h1:7X/raswPFr05uY3HiLlYeyQntB6OO7E/d2Cu7qoaN2w=
-github.com/jsimonetti/rtnetlink/v2 v2.0.1 h1:xda7qaHDSVOsADNouv7ukSuicKZO7GgVUCXxpaIEIlM=
-github.com/jsimonetti/rtnetlink/v2 v2.0.1/go.mod h1:7MoNYNbb3UaDHtF8udiJo/RH6VsTKP1pqKLUTVCvToE=
-github.com/kr/pretty v0.3.1 h1:flRD4NNwYAUpkphVc1HcthR4KEIFJ65n8Mw5qdRn3LE=
-github.com/kr/pretty v0.3.1/go.mod h1:hoEshYVHaxMs3cyo3Yncou5ZscifuDolrwPKZanG3xk=
-github.com/kr/text v0.2.0 h1:5Nx0Ya0ZqY2ygV366QzturHI13Jq95ApcVaJBhpS+AY=
-github.com/kr/text v0.2.0/go.mod h1:eLer722TekiGuMkidMxC/pM04lWEeraHUUmBw8l2grE=
 github.com/lufia/plan9stats v0.0.0-20211012122336-39d0f177ccd0 h1:6E+4a0GO5zZEnZ81pIr0yLvtUWk2if982qA3F3QD6H4=
 github.com/lufia/plan9stats v0.0.0-20211012122336-39d0f177ccd0/go.mod h1:zJYVVT2jmtg6P3p1VtQj7WsuWi/y4VnjVBn7F8KPB3I=
-github.com/mdlayher/netlink v1.7.2 h1:/UtM3ofJap7Vl4QWCPDGXY8d3GIY2UGSDbK+QWmY8/g=
-github.com/mdlayher/netlink v1.7.2/go.mod h1:xraEF7uJbxLhc5fpHL4cPe221LI2bdttWlU+ZGLfQSw=
-github.com/mdlayher/socket v0.4.1 h1:eM9y2/jlbs1M615oshPQOHZzj6R6wMT7bX5NPiQvn2U=
-github.com/mdlayher/socket v0.4.1/go.mod h1:cAqeGjoufqdxWkD7DkpyS+wcefOtmu5OQ8KuoJGIReA=
 github.com/pmezard/go-difflib v1.0.0 h1:4DBwDE0NGyQoBHbLQYPwSUPoCMWR5BEzIk/f1lZbAQM=
 github.com/pmezard/go-difflib v1.0.0/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZNVY4sRDYZ/4=
 github.com/power-devops/perfstat v0.0.0-20210106213030-5aafc221ea8c h1:ncq/mPwQF4JjgDlrVEn3C11VoGHZN7m8qihwgMEtzYw=
 github.com/power-devops/perfstat v0.0.0-20210106213030-5aafc221ea8c/go.mod h1:OmDBASR4679mdNQnz2pUhc2G8CO2JrUAVFDRBDP/hJE=
-github.com/rogpeppe/go-internal v1.12.0 h1:exVL4IDcn6na9z1rAb56Vxr+CgyK3nn3O+epU5NdKM8=
-github.com/rogpeppe/go-internal v1.12.0/go.mod h1:E+RYuTGaKKdloAfM02xzb0FW3Paa99yedzYV+kq4uf4=
 github.com/sashabaranov/go-openai v1.32.0 h1:Yk3iE9moX3RBXxrof3OBtUBrE7qZR0zF9ebsoO4zVzI=
 github.com/sashabaranov/go-openai v1.32.0/go.mod h1:lj5b/K+zjTSFxVLijLSTDZuP7adOgerWeFyZLUhAKRg=
 github.com/shirou/gopsutil/v3 v3.24.5 h1:i0t8kL+kQTvpAYToeuiVk3TgDeKOFioZO3Ztz/iZ9pI=
@@ -49,10 +31,6 @@ github.com/tklauser/numcpus v0.6.1 h1:ng9scYS7az0Bk4OZLvrNXNSAO2Pxr1XXRAPyjhIx+F
 github.com/tklauser/numcpus v0.6.1/go.mod h1:1XfjsgE2zo8GVw7POkMbHENHzVg3GzmoZ9fESEdAacY=
 github.com/yusufpapurcu/wmi v1.2.4 h1:zFUKzehAFReQwLys1b/iSMl+JQGSCSjtVqQn9bBrPo0=
 github.com/yusufpapurcu/wmi v1.2.4/go.mod h1:SBZ9tNy3G9/m5Oi98Zks0QjeHVDvuK0qfxQmPyzfmi0=
-golang.org/x/net v0.38.0 h1:vRMAPTMaeGqVhG5QyLJHqNDwecKTomGeqbnfZyKlBI8=
-golang.org/x/net v0.38.0/go.mod h1:ivrbrMbzFq5J41QOQh0siUuly180yBYtLp+CKbEaFx8=
-golang.org/x/sync v0.1.0 h1:wsuoTGHzEhffawBOhz5CYhcrV4IdKZbEyZjBMuTp12o=
-golang.org/x/sync v0.1.0/go.mod h1:RxMgew5VJxzue5/jJTE5uejpjVlOe/izrB70Jof72aM=
 golang.org/x/sys v0.0.0-20190916202348-b4ddaad3f8a3/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20201204225414-ed752295db88/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.8.0/go.mod h1:oPkhp1MJrh7nUepCBck5+mAzfO9JrbApNNgaTdGDITg=

internal/auth/auth.go

@@ -13,6 +13,7 @@ import (
 	"time"
 
 	"nannyagentv2/internal/config"
+	"nannyagentv2/internal/logging"
 	"nannyagentv2/internal/types"
 )
 
@@ -103,7 +104,7 @@ func (am *AuthManager) StartDeviceAuthorization() (*types.DeviceAuthResponse, er
 
 // PollForToken polls the token endpoint until authorization is complete
 func (am *AuthManager) PollForToken(deviceCode string) (*types.TokenResponse, error) {
-	fmt.Println("⏳ Waiting for user authorization...")
+	logging.Info("Waiting for user authorization...")
 
 	for attempts := 0; attempts < MaxPollAttempts; attempts++ {
 		tokenReq := types.TokenRequest{
@@ -151,7 +152,7 @@ func (am *AuthManager) PollForToken(deviceCode string) (*types.TokenResponse, er
 		}
 
 		if tokenResp.AccessToken != "" {
-			fmt.Println("\n✅ Authorization successful!")
+			logging.Info("Authorization successful!")
 			return &tokenResp, nil
 		}
 
@@ -230,7 +231,7 @@ func (am *AuthManager) SaveToken(token *types.AuthToken) error {
 		refreshTokenPath := filepath.Join(TokenStorageDir, RefreshTokenFile)
 		if err := os.WriteFile(refreshTokenPath, []byte(token.RefreshToken), 0600); err != nil {
 			// Don't fail if refresh token backup fails, just log
-			fmt.Printf("Warning: Failed to save backup refresh token: %v\n", err)
+			logging.Warning("Failed to save backup refresh token: %v", err)
 		}
 	}
 
@@ -271,8 +272,8 @@ func (am *AuthManager) RegisterDevice() (*types.AuthToken, error) {
 		return nil, fmt.Errorf("failed to start device authorization: %w", err)
 	}
 
-	fmt.Printf("Please visit: %s\n", deviceAuth.VerificationURI)
-	fmt.Printf("And enter code: %s\n", deviceAuth.UserCode)
+	logging.Info("Please visit: %s", deviceAuth.VerificationURI)
+	logging.Info("And enter code: %s", deviceAuth.UserCode)
 
 	// Step 2: Poll for token
 	tokenResp, err := am.PollForToken(deviceAuth.DeviceCode)
@@ -318,13 +319,13 @@ func (am *AuthManager) EnsureAuthenticated() (*types.AuthToken, error) {
 			// Try to load refresh token from backup file
 			if backupRefreshToken, backupErr := am.loadRefreshTokenFromBackup(); backupErr == nil {
 				refreshToken = backupRefreshToken
-				fmt.Println("🔄 Found backup refresh token, attempting to use it...")
+				logging.Debug("Found backup refresh token, attempting to use it...")
 			}
 		}
 	}
 
 	if refreshToken != "" {
-		fmt.Println("🔄 Attempting to refresh access token...")
+		logging.Debug("Attempting to refresh access token...")
 
 		refreshResp, refreshErr := am.RefreshAccessToken(refreshToken)
 		if refreshErr == nil {

internal/config/config.go

@@ -6,6 +6,8 @@ import (
 	"path/filepath"
 	"strings"
 
+	"nannyagentv2/internal/logging"
+
 	"github.com/joho/godotenv"
 )
 
@@ -39,9 +41,9 @@ func LoadConfig() (*Config, error) {
 	envFile := findEnvFile()
 	if envFile != "" {
 		if err := godotenv.Load(envFile); err != nil {
-			fmt.Printf("Warning: Could not load .env file from %s: %v\n", envFile, err)
+			logging.Warning("Could not load .env file from %s: %v", envFile, err)
 		} else {
-			fmt.Printf("Loaded configuration from %s\n", envFile)
+			logging.Info("Loaded configuration from %s", envFile)
 		}
 	}
 
@@ -124,8 +126,8 @@ func (c *Config) PrintConfig() {
 		return
 	}
 
-	fmt.Println("Configuration:")
-	fmt.Printf("  Supabase Project URL: %s\n", c.SupabaseProjectURL)
-	fmt.Printf("  Metrics Interval: %d seconds\n", c.MetricsInterval)
-	fmt.Printf("  Debug: %v\n", c.Debug)
+	logging.Debug("Configuration:")
+	logging.Debug("  Supabase Project URL: %s", c.SupabaseProjectURL)
+	logging.Debug("  Metrics Interval: %d seconds", c.MetricsInterval)
+	logging.Debug("  Debug: %v", c.Debug)
 }

internal/logging/logger.go

@@ -5,11 +5,39 @@ import (
 	"log"
 	"log/syslog"
 	"os"
+	"strings"
 )
 
+// LogLevel defines the logging level
+type LogLevel int
+
+const (
+	LevelDebug LogLevel = iota
+	LevelInfo
+	LevelWarning
+	LevelError
+)
+
+func (l LogLevel) String() string {
+	switch l {
+	case LevelDebug:
+		return "DEBUG"
+	case LevelInfo:
+		return "INFO"
+	case LevelWarning:
+		return "WARN"
+	case LevelError:
+		return "ERROR"
+	default:
+		return "INFO"
+	}
+}
+
+// Logger provides structured logging with configurable levels
 type Logger struct {
 	syslogWriter *syslog.Writer
-	debugMode    bool
+	level        LogLevel
+	showEmoji    bool
 }
 
 var defaultLogger *Logger
@@ -18,9 +46,16 @@ func init() {
 	defaultLogger = NewLogger()
 }
 
+// NewLogger creates a new logger with default configuration
 func NewLogger() *Logger {
+	return NewLoggerWithLevel(getLogLevelFromEnv())
+}
+
+// NewLoggerWithLevel creates a logger with specified level
+func NewLoggerWithLevel(level LogLevel) *Logger {
 	l := &Logger{
-		debugMode: os.Getenv("DEBUG") == "true",
+		level:     level,
+		showEmoji: os.Getenv("LOG_NO_EMOJI") != "true",
 	}
 
 	// Try to connect to syslog
@@ -31,39 +66,87 @@ func NewLogger() *Logger {
 	return l
 }
 
-func (l *Logger) Info(format string, args ...interface{}) {
-	msg := fmt.Sprintf(format, args...)
-	if l.syslogWriter != nil {
-		l.syslogWriter.Info(msg)
+// getLogLevelFromEnv parses log level from environment variable
+func getLogLevelFromEnv() LogLevel {
+	level := strings.ToUpper(os.Getenv("LOG_LEVEL"))
+	switch level {
+	case "DEBUG":
+		return LevelDebug
+	case "INFO", "":
+		return LevelInfo
+	case "WARN", "WARNING":
+		return LevelWarning
+	case "ERROR":
+		return LevelError
+	default:
+		return LevelInfo
 	}
-	log.Printf("[INFO] %s", msg)
+}
+
+// logMessage handles the actual logging
+func (l *Logger) logMessage(level LogLevel, format string, args ...interface{}) {
+	if level < l.level {
+		return
+	}
+
+	msg := fmt.Sprintf(format, args...)
+	prefix := fmt.Sprintf("[%s]", level.String())
+
+	// Add emoji prefix if enabled
+	if l.showEmoji {
+		switch level {
+		case LevelDebug:
+			prefix = "🔍 " + prefix
+		case LevelInfo:
+			prefix = "ℹ️  " + prefix
+		case LevelWarning:
+			prefix = "⚠️  " + prefix
+		case LevelError:
+			prefix = "❌ " + prefix
+		}
+	}
+
+	// Log to syslog if available
+	if l.syslogWriter != nil {
+		switch level {
+		case LevelDebug:
+			l.syslogWriter.Debug(msg)
+		case LevelInfo:
+			l.syslogWriter.Info(msg)
+		case LevelWarning:
+			l.syslogWriter.Warning(msg)
+		case LevelError:
+			l.syslogWriter.Err(msg)
+		}
+	}
+
+	log.Printf("%s %s", prefix, msg)
 }
 
 func (l *Logger) Debug(format string, args ...interface{}) {
-	if !l.debugMode {
-		return
+	l.logMessage(LevelDebug, format, args...)
 }
-	msg := fmt.Sprintf(format, args...)
-	if l.syslogWriter != nil {
-		l.syslogWriter.Debug(msg)
-	}
-	log.Printf("[DEBUG] %s", msg)
+
+func (l *Logger) Info(format string, args ...interface{}) {
+	l.logMessage(LevelInfo, format, args...)
 }
 
 func (l *Logger) Warning(format string, args ...interface{}) {
-	msg := fmt.Sprintf(format, args...)
-	if l.syslogWriter != nil {
-		l.syslogWriter.Warning(msg)
-	}
-	log.Printf("[WARNING] %s", msg)
+	l.logMessage(LevelWarning, format, args...)
 }
 
 func (l *Logger) Error(format string, args ...interface{}) {
-	msg := fmt.Sprintf(format, args...)
-	if l.syslogWriter != nil {
-		l.syslogWriter.Err(msg)
+	l.logMessage(LevelError, format, args...)
 }
-	log.Printf("[ERROR] %s", msg)
+
+// SetLevel changes the logging level
+func (l *Logger) SetLevel(level LogLevel) {
+	l.level = level
+}
+
+// GetLevel returns current logging level
+func (l *Logger) GetLevel() LogLevel {
+	return l.level
 }
 
 func (l *Logger) Close() {
@@ -73,14 +156,14 @@ func (l *Logger) Close() {
 }
 
 // Global logging functions
-func Info(format string, args ...interface{}) {
-	defaultLogger.Info(format, args...)
-}
-
 func Debug(format string, args ...interface{}) {
 	defaultLogger.Debug(format, args...)
 }
 
+func Info(format string, args ...interface{}) {
+	defaultLogger.Info(format, args...)
+}
+
 func Warning(format string, args ...interface{}) {
 	defaultLogger.Warning(format, args...)
 }
@@ -88,3 +171,13 @@ func Warning(format string, args ...interface{}) {
 func Error(format string, args ...interface{}) {
 	defaultLogger.Error(format, args...)
 }
+
+// SetLevel sets the global logger level
+func SetLevel(level LogLevel) {
+	defaultLogger.SetLevel(level)
+}
+
+// GetLevel gets the global logger level
+func GetLevel() LogLevel {
+	return defaultLogger.GetLevel()
+}

internal/types/types.go

@@ -241,7 +241,7 @@ type CommandResult struct {
 type EBPFEnhancedDiagnosticResponse struct {
 	ResponseType string        `json:"response_type"`
 	Reasoning    string        `json:"reasoning"`
-	Commands     []Command     `json:"commands"`
+	Commands     []string      `json:"commands"` // Changed to []string to match current prompt format
 	EBPFPrograms []EBPFRequest `json:"ebpf_programs"`
 	NextActions  []string      `json:"next_actions,omitempty"`
 }

investigation_server.go

@@ -9,6 +9,7 @@ import (
 	"time"
 
 	"nannyagentv2/internal/auth"
+	"nannyagentv2/internal/logging"
 	"nannyagentv2/internal/metrics"
 
 	"github.com/sashabaranov/go-openai"
@@ -62,7 +63,7 @@ func NewInvestigationServer(agent *LinuxDiagnosticAgent, authManager *auth.AuthM
 			agentID = id
 
 		} else {
-			fmt.Printf("❌ Failed to get agent ID from auth manager: %v\n", err)
+			logging.Error("Failed to get agent ID from auth manager: %v", err)
 		}
 	}
 
@@ -117,9 +118,9 @@ func (s *InvestigationServer) Start() error {
 	// Start realtime polling for backend-initiated investigations
 	if s.supabaseURL != "" && s.authManager != nil {
 		go s.startRealtimePolling()
-		fmt.Printf("🔄 Realtime investigation polling enabled\n")
+		logging.Info("Realtime investigation polling enabled")
 	} else {
-		fmt.Printf("⚠️  Realtime investigation polling disabled (missing Supabase config or auth)\n")
+		logging.Warning("Realtime investigation polling disabled (missing Supabase config or auth)")
 	}
 
 	server := &http.Server{
@@ -129,7 +130,7 @@ func (s *InvestigationServer) Start() error {
 		WriteTimeout: 30 * time.Second,
 	}
 
-	fmt.Printf("🔍 Investigation server started on port %s (Agent ID: %s)\n", s.port, s.agentID)
+	logging.Info("Investigation server started on port %s (Agent ID: %s)", s.port, s.agentID)
 	return server.ListenAndServe()
 }
 
@@ -221,7 +222,7 @@ func (s *InvestigationServer) sendCommandResultsToTensorZero(diagnosticResp Diag
 	})
 
 	// Send to TensorZero via application agent's sendRequest method
-	fmt.Printf("🔄 Sending command results to TensorZero for analysis...\n")
+	logging.Debug("Sending command results to TensorZero for analysis")
 	response, err := s.applicationAgent.sendRequest(messages)
 	if err != nil {
 		return nil, fmt.Errorf("failed to send request to TensorZero: %w", err)
@@ -232,7 +233,7 @@ func (s *InvestigationServer) sendCommandResultsToTensorZero(diagnosticResp Diag
 	}
 
 	content := response.Choices[0].Message.Content
-	fmt.Printf("🤖 TensorZero continued analysis:\n%s\n", content)
+	logging.Debug("TensorZero continued analysis: %s", content)
 
 	// Try to parse the response to determine if it's diagnostic or resolution
 	var diagnosticNextResp DiagnosticResponse
@@ -240,7 +241,7 @@ func (s *InvestigationServer) sendCommandResultsToTensorZero(diagnosticResp Diag
 
 	// Check if it's another diagnostic response
 	if err := json.Unmarshal([]byte(content), &diagnosticNextResp); err == nil && diagnosticNextResp.ResponseType == "diagnostic" {
-		fmt.Printf("🔄 TensorZero requests %d more commands\n", len(diagnosticNextResp.Commands))
+		logging.Debug("TensorZero requests %d more commands", len(diagnosticNextResp.Commands))
 		return map[string]interface{}{
 			"type":     "diagnostic",
 			"response": diagnosticNextResp,
@@ -295,7 +296,7 @@ func (s *InvestigationServer) handleInvestigation(w http.ResponseWriter, r *http
 		return
 	}
 
-	fmt.Printf("📋 Received investigation payload with response_type: %s\n", responseType)
+	logging.Debug("Received investigation payload with response_type: %s", responseType)
 
 	switch responseType {
 	case "diagnostic":

main.go

@@ -13,6 +13,7 @@ import (
 
 	"nannyagentv2/internal/auth"
 	"nannyagentv2/internal/config"
+	"nannyagentv2/internal/logging"
 	"nannyagentv2/internal/metrics"
 	"nannyagentv2/internal/types"
 )
@@ -22,12 +23,9 @@ const Version = "v2.0.0"
 // checkRootPrivileges ensures the program is running as root
 func checkRootPrivileges() {
 	if os.Geteuid() != 0 {
-		fmt.Fprintf(os.Stderr, "❌ ERROR: This program must be run as root for eBPF functionality.\n")
-		fmt.Fprintf(os.Stderr, "Please run with: sudo %s\n", os.Args[0])
-		fmt.Fprintf(os.Stderr, "Reason: eBPF programs require root privileges to:\n")
-		fmt.Fprintf(os.Stderr, " - Load programs into the kernel\n")
-		fmt.Fprintf(os.Stderr, " - Attach to kernel functions and tracepoints\n")
-		fmt.Fprintf(os.Stderr, " - Access kernel memory maps\n")
+		logging.Error("This program must be run as root for eBPF functionality")
+		logging.Error("Please run with: sudo %s", os.Args[0])
+		logging.Error("Reason: eBPF programs require root privileges to:\n - Load programs into the kernel\n - Attach to kernel functions and tracepoints\n - Access kernel memory maps")
 		os.Exit(1)
 	}
 }
@@ -36,7 +34,7 @@ func checkRootPrivileges() {
 func checkKernelVersionCompatibility() {
 	output, err := exec.Command("uname", "-r").Output()
 	if err != nil {
-		fmt.Fprintf(os.Stderr, "❌ ERROR: Cannot determine kernel version: %v\n", err)
+		logging.Error("Cannot determine kernel version: %v", err)
 		os.Exit(1)
 	}
 
@@ -45,66 +43,57 @@ func checkKernelVersionCompatibility() {
 	// Parse version (e.g., "5.15.0-56-generic" -> major=5, minor=15)
 	parts := strings.Split(kernelVersion, ".")
 	if len(parts) < 2 {
-		fmt.Fprintf(os.Stderr, "❌ ERROR: Cannot parse kernel version: %s\n", kernelVersion)
+		logging.Error("Cannot parse kernel version: %s", kernelVersion)
 		os.Exit(1)
 	}
 
 	major, err := strconv.Atoi(parts[0])
 	if err != nil {
-		fmt.Fprintf(os.Stderr, "❌ ERROR: Cannot parse major kernel version: %s\n", parts[0])
+		logging.Error("Cannot parse major kernel version: %s", parts[0])
 		os.Exit(1)
 	}
 
 	minor, err := strconv.Atoi(parts[1])
 	if err != nil {
-		fmt.Fprintf(os.Stderr, "❌ ERROR: Cannot parse minor kernel version: %s\n", parts[1])
+		logging.Error("Cannot parse minor kernel version: %s", parts[1])
 		os.Exit(1)
 	}
 
 	// Check if kernel is 4.4 or higher
 	if major < 4 || (major == 4 && minor < 4) {
-		fmt.Fprintf(os.Stderr, "❌ ERROR: Kernel version %s is too old for eBPF.\n", kernelVersion)
-		fmt.Fprintf(os.Stderr, "Required: Linux kernel 4.4 or higher\n")
-		fmt.Fprintf(os.Stderr, "Current: %s\n", kernelVersion)
-		fmt.Fprintf(os.Stderr, "Reason: eBPF requires kernel features introduced in 4.4+:\n")
-		fmt.Fprintf(os.Stderr, " - BPF system call support\n")
-		fmt.Fprintf(os.Stderr, " - eBPF program types (kprobe, tracepoint)\n")
-		fmt.Fprintf(os.Stderr, " - BPF maps and helper functions\n")
+		logging.Error("Kernel version %s is too old for eBPF", kernelVersion)
+		logging.Error("Required: Linux kernel 4.4 or higher")
+		logging.Error("Current: %s", kernelVersion)
+		logging.Error("Reason: eBPF requires kernel features introduced in 4.4+:\n - BPF system call support\n - eBPF program types (kprobe, tracepoint)\n - BPF maps and helper functions")
 		os.Exit(1)
 	}
-
 }
 
 // checkEBPFSupport validates eBPF subsystem availability
 func checkEBPFSupport() {
 	// Check if /sys/kernel/debug/tracing exists (debugfs mounted)
 	if _, err := os.Stat("/sys/kernel/debug/tracing"); os.IsNotExist(err) {
-		fmt.Fprintf(os.Stderr, "⚠️ WARNING: debugfs not mounted. Some eBPF features may not work.\n")
-		fmt.Fprintf(os.Stderr, "To fix: sudo mount -t debugfs debugfs /sys/kernel/debug\n")
+		logging.Warning("debugfs not mounted. Some eBPF features may not work")
+		logging.Info("To fix: sudo mount -t debugfs debugfs /sys/kernel/debug")
 	}
 
 	// Check if we can access BPF syscall
 	fd, _, errno := syscall.Syscall(321, 0, 0, 0) // BPF syscall number on x86_64
 	if errno != 0 && errno != syscall.EINVAL {
-		fmt.Fprintf(os.Stderr, "❌ ERROR: BPF syscall not available (errno: %v)\n", errno)
-		fmt.Fprintf(os.Stderr, "This may indicate:\n")
-		fmt.Fprintf(os.Stderr, " - Kernel compiled without BPF support\n")
-		fmt.Fprintf(os.Stderr, " - BPF syscall disabled in kernel config\n")
+		logging.Error("BPF syscall not available (errno: %v)", errno)
+		logging.Error("This may indicate:\n - Kernel compiled without BPF support\n - BPF syscall disabled in kernel config")
 		os.Exit(1)
 	}
 	if fd > 0 {
 		syscall.Close(int(fd))
 	}
-
 }
 
 // runInteractiveDiagnostics starts the interactive diagnostic session
 func runInteractiveDiagnostics(agent *LinuxDiagnosticAgent) {
-	fmt.Println("")
-	fmt.Println("🔍 Linux eBPF-Enhanced Diagnostic Agent")
-	fmt.Println("=======================================")
-	fmt.Println("Linux Diagnostic Agent Started")
-	fmt.Println("Enter a system issue description (or 'quit' to exit):")
+	logging.Info("=== Linux eBPF-Enhanced Diagnostic Agent ===")
+	logging.Info("Linux Diagnostic Agent Started")
+	logging.Info("Enter a system issue description (or 'quit' to exit):")
 
 	scanner := bufio.NewScanner(os.Stdin)
 	for {
@@ -124,7 +113,7 @@ func runInteractiveDiagnostics(agent *LinuxDiagnosticAgent) {
 
 		// Process the issue with AI capabilities via TensorZero
 		if err := agent.DiagnoseIssue(input); err != nil {
-			fmt.Printf("Error: %v\n", err)
+			logging.Error("Diagnosis failed: %v", err)
 		}
 	}
 
@@ -132,19 +121,18 @@ func runInteractiveDiagnostics(agent *LinuxDiagnosticAgent) {
 		log.Fatal(err)
 	}
 
-	fmt.Println("Goodbye!")
+	logging.Info("Goodbye!")
 }
 
 func main() {
-	fmt.Printf("🚀 NannyAgent v%s starting...\n", Version)
+	logging.Info("NannyAgent v%s starting...", Version)
 
 	// Perform system compatibility checks first
-	fmt.Println("Performing system compatibility checks...")
+	logging.Info("Performing system compatibility checks...")
 	checkRootPrivileges()
 	checkKernelVersionCompatibility()
 	checkEBPFSupport()
-	fmt.Println("✅ All system checks passed")
-	fmt.Println("")
+	logging.Info("All system checks passed")
 
 	// Load configuration
 	cfg, err := config.LoadConfig()
@@ -164,10 +152,10 @@ func main() {
 		log.Fatalf("❌ Authentication failed: %v", err)
 	}
 
-	fmt.Println("✅ Authentication successful!")
+	logging.Info("Authentication successful!")
 
-	// Initialize the diagnostic agent for interactive CLI use
-	agent := NewLinuxDiagnosticAgent()
+	// Initialize the diagnostic agent for interactive CLI use with authentication
+	agent := NewLinuxDiagnosticAgentWithAuth(authManager)
 
 	// Initialize a separate agent for WebSocket investigations using the application model
 	applicationAgent := NewLinuxDiagnosticAgent()
@@ -177,53 +165,53 @@ func main() {
 	wsClient := NewWebSocketClient(applicationAgent, authManager)
 	go func() {
 		if err := wsClient.Start(); err != nil {
-			log.Printf("❌ WebSocket client error: %v", err)
+			logging.Error("WebSocket client error: %v", err)
 		}
 	}()
 
 	// Start background metrics collection in a goroutine
 	go func() {
-		fmt.Println("❤️  Starting background metrics collection and heartbeat...")
+		logging.Debug("Starting background metrics collection and heartbeat...")
 
 		ticker := time.NewTicker(time.Duration(cfg.MetricsInterval) * time.Second)
 		defer ticker.Stop()
 
 		// Send initial heartbeat
 		if err := sendHeartbeat(cfg, token, metricsCollector); err != nil {
-			log.Printf("⚠️  Initial heartbeat failed: %v", err)
+			logging.Warning("Initial heartbeat failed: %v", err)
 		}
 
 		// Main heartbeat loop
 		for range ticker.C {
 			// Check if token needs refresh
 			if authManager.IsTokenExpired(token) {
-				fmt.Println("🔄 Token expiring soon, refreshing...")
+				logging.Debug("Token expiring soon, refreshing...")
 				newToken, refreshErr := authManager.EnsureAuthenticated()
 				if refreshErr != nil {
-					log.Printf("❌ Token refresh failed: %v", refreshErr)
+					logging.Warning("Token refresh failed: %v", refreshErr)
 					continue
 				}
 				token = newToken
-				fmt.Println("✅ Token refreshed successfully")
+				logging.Debug("Token refreshed successfully")
 			}
 
 			// Send heartbeat
 			if err := sendHeartbeat(cfg, token, metricsCollector); err != nil {
-				log.Printf("⚠️  Heartbeat failed: %v", err)
+				logging.Warning("Heartbeat failed: %v", err)
 
 				// If unauthorized, try to refresh token
 				if err.Error() == "unauthorized" {
-					fmt.Println("🔄 Unauthorized, attempting token refresh...")
+					logging.Debug("Unauthorized, attempting token refresh...")
 					newToken, refreshErr := authManager.EnsureAuthenticated()
 					if refreshErr != nil {
-						log.Printf("❌ Token refresh failed: %v", refreshErr)
+						logging.Warning("Token refresh failed: %v", refreshErr)
 						continue
 					}
 					token = newToken
 
 					// Retry heartbeat with new token (silently)
 					if retryErr := sendHeartbeat(cfg, token, metricsCollector); retryErr != nil {
-						log.Printf("⚠️  Retry heartbeat failed: %v", retryErr)
+						logging.Warning("Retry heartbeat failed: %v", retryErr)
 					}
 				}
 			}

scripts/debug_trace_script.sh

@@ -0,0 +1,19 @@
+#!/bin/bash
+
+# Test the current script generation
+echo "Testing tracepoint script generation..."
+
+# Simulate what the failing test does
+echo "Target: syscalls:sys_enter_openat"
+echo "ProbeType: t"
+echo ""
+echo "Generated bpftrace script would be:"
+echo "tracepoint:syscalls:sys_enter_openat {"
+echo "  printf(\"TRACE|%d|%d|%d|%s|syscalls:sys_enter_openat|file access\\n\", nsecs, pid, tid, comm, arg2@user);"
+echo "}"
+echo ""
+echo "This is INVALID - should be:"
+echo "tracepoint:syscalls:sys_enter_openat {"
+echo "  printf(\"TRACE|%d|%d|%d|%s|openat|file access\\n\", nsecs, pid, tid, comm);"
+echo "}"
+

websocket_client.go

@@ -13,7 +13,9 @@ import (
 	"time"
 
 	"nannyagentv2/internal/auth"
+	"nannyagentv2/internal/logging"
 	"nannyagentv2/internal/metrics"
+	"nannyagentv2/internal/types"
 
 	"github.com/gorilla/websocket"
 	"github.com/sashabaranov/go-openai"
@@ -74,7 +76,7 @@ func NewWebSocketClient(agent *LinuxDiagnosticAgent, authManager *auth.AuthManag
 			agentID = id
 			// Agent ID retrieved successfully
 		} else {
-			fmt.Printf("❌ Failed to get agent ID from auth manager: %v\n", err)
+			logging.Error("Failed to get agent ID from auth manager: %v", err)
 		}
 	}
 
@@ -178,7 +180,7 @@ func (c *WebSocketClient) connect() error {
 	if err != nil {
 		c.consecutiveFailures++
 		if c.consecutiveFailures >= 5 && resp != nil {
-			fmt.Printf("❌ WebSocket handshake failed with status: %d (failure #%d)\n", resp.StatusCode, c.consecutiveFailures)
+			logging.Error("WebSocket handshake failed with status: %d (failure #%d)", resp.StatusCode, c.consecutiveFailures)
 		}
 		return fmt.Errorf("websocket connection failed: %v", err)
 	}
@@ -205,7 +207,7 @@ func (c *WebSocketClient) handleMessages() {
 		case <-c.ctx.Done():
 			// Only log context cancellation if there have been failures
 			if c.consecutiveFailures >= 5 {
-				fmt.Printf("📡 Context cancelled after %v, stopping message handler\n", time.Since(connectionStart))
+				logging.Debug("Context cancelled after %v, stopping message handler", time.Since(connectionStart))
 			}
 			return
 		default:
@@ -223,14 +225,14 @@ func (c *WebSocketClient) handleMessages() {
 				// Only log specific errors after failure threshold
 				if c.consecutiveFailures >= 5 {
 					if websocket.IsCloseError(err, websocket.CloseNormalClosure, websocket.CloseGoingAway) {
-						log.Printf("🔒 WebSocket closed normally after %v: %v", connectionDuration, err)
+						logging.Debug("WebSocket closed normally after %v: %v", connectionDuration, err)
 					} else if websocket.IsUnexpectedCloseError(err, websocket.CloseGoingAway, websocket.CloseAbnormalClosure) {
-						log.Printf("💥 ABNORMAL CLOSE after %v (code 1006 = server-side timeout/kill): %v", connectionDuration, err)
-						log.Printf("🕒 Last read took %v, connection lived %v", readDuration, connectionDuration)
+						logging.Error("ABNORMAL CLOSE after %v (code 1006 = server-side timeout/kill): %v", connectionDuration, err)
+						logging.Debug("Last read took %v, connection lived %v", readDuration, connectionDuration)
 					} else if netErr, ok := err.(net.Error); ok && netErr.Timeout() {
-						log.Printf("⏰ READ TIMEOUT after %v: %v", connectionDuration, err)
+						logging.Warning("READ TIMEOUT after %v: %v", connectionDuration, err)
 					} else {
-						log.Printf("❌ WebSocket error after %v: %v", connectionDuration, err)
+						logging.Error("WebSocket error after %v: %v", connectionDuration, err)
 					}
 				}
 
@@ -239,7 +241,7 @@ func (c *WebSocketClient) handleMessages() {
 
 				// Only show diagnostics after multiple failures
 				if c.consecutiveFailures >= 5 {
-					log.Printf("🔍 DIAGNOSTIC - Connection failed #%d after %v", c.consecutiveFailures, connectionDuration)
+					logging.Debug("DIAGNOSTIC - Connection failed #%d after %v", c.consecutiveFailures, connectionDuration)
 				}
 
 				// Attempt reconnection instead of returning immediately
@@ -265,7 +267,7 @@ func (c *WebSocketClient) handleMessages() {
 				// Task result acknowledged
 
 			default:
-				log.Printf("⚠️  Unknown message type: %s", message.Type)
+				logging.Warning("Unknown message type: %s", message.Type)
 			}
 		}
 	}
@@ -276,14 +278,14 @@ func (c *WebSocketClient) handleInvestigationTask(data interface{}) {
 	// Parse task data
 	taskBytes, err := json.Marshal(data)
 	if err != nil {
-		log.Printf("❌ Error marshaling task data: %v", err)
+		logging.Error("Error marshaling task data: %v", err)
 		return
 	}
 
 	var task InvestigationTask
 	err = json.Unmarshal(taskBytes, &task)
 	if err != nil {
-		log.Printf("❌ Error unmarshaling investigation task: %v", err)
+		logging.Error("Error unmarshaling investigation task: %v", err)
 		return
 	}
 
@@ -300,7 +302,7 @@ func (c *WebSocketClient) handleInvestigationTask(data interface{}) {
 
 	if err != nil {
 		taskResult.Error = err.Error()
-		fmt.Printf("❌ Task execution failed: %v\n", err)
+		logging.Error("Task execution failed: %v", err)
 	} else {
 		taskResult.CommandResults = results
 		// Task executed successfully
@@ -356,7 +358,7 @@ func (c *WebSocketClient) executeDiagnosticCommands(diagnosticPayload map[string
 
 		if err != nil {
 			result["error"] = err.Error()
-			fmt.Printf("❌ Command [%s] failed: %v (exit code: %d)\n", id, err, exitCode)
+			logging.Warning("Command [%s] failed: %v (exit code: %d)", id, err, exitCode)
 		}
 
 		commandResults = append(commandResults, result)
@@ -379,7 +381,7 @@ func (c *WebSocketClient) executeDiagnosticCommands(diagnosticPayload map[string
 
 // executeEBPFPrograms executes eBPF monitoring programs using the real eBPF manager
 func (c *WebSocketClient) executeEBPFPrograms(ebpfPrograms []interface{}) []map[string]interface{} {
-	var ebpfRequests []EBPFRequest
+	var ebpfRequests []types.EBPFRequest
 
 	// Convert interface{} to EBPFRequest structs
 	for _, prog := range ebpfPrograms {
@@ -398,7 +400,7 @@ func (c *WebSocketClient) executeEBPFPrograms(ebpfPrograms []interface{}) []map[
 			continue
 		}
 
-		ebpfRequests = append(ebpfRequests, EBPFRequest{
+		ebpfRequests = append(ebpfRequests, types.EBPFRequest{
 			Name:        name,
 			Type:        progType,
 			Target:      target,
@@ -444,7 +446,7 @@ func (c *WebSocketClient) executeCommandsFromPayload(commands []interface{}) []m
 
 		if err != nil {
 			result["error"] = err.Error()
-			fmt.Printf("❌ Command [%s] failed: %v (exit code: %d)\n", id, err, exitCode)
+			logging.Warning("Command [%s] failed: %v (exit code: %d)", id, err, exitCode)
 		}
 
 		commandResults = append(commandResults, result)
@@ -502,7 +504,7 @@ func (c *WebSocketClient) sendTaskResult(result TaskResult) {
 
 	err := c.conn.WriteJSON(message)
 	if err != nil {
-		log.Printf("❌ Error sending task result: %v", err)
+		logging.Error("Error sending task result: %v", err)
 	}
 }
 
@@ -516,7 +518,7 @@ func (c *WebSocketClient) startHeartbeat() {
 	for {
 		select {
 		case <-c.ctx.Done():
-			fmt.Printf("💓 Heartbeat stopped due to context cancellation\n")
+			logging.Debug("Heartbeat stopped due to context cancellation")
 			return
 		case <-ticker.C:
 			// Sending heartbeat
@@ -531,8 +533,8 @@ func (c *WebSocketClient) startHeartbeat() {
 
 			err := c.conn.WriteJSON(heartbeat)
 			if err != nil {
-				log.Printf("❌ Error sending heartbeat: %v", err)
-				fmt.Printf("💓 Heartbeat failed, connection likely dead\n")
+				logging.Error("Error sending heartbeat: %v", err)
+				logging.Debug("Heartbeat failed, connection likely dead")
 				return
 			}
 			// Heartbeat sent
@@ -656,14 +658,14 @@ func (c *WebSocketClient) handlePendingInvestigation(investigation PendingInvest
 	for {
 		tzResp, tzErr := c.agent.sendRequestWithEpisode(messages, episodeID)
 		if tzErr != nil {
-			fmt.Printf("⚠️  TensorZero continuation failed: %v\n", tzErr)
+			logging.Warning("TensorZero continuation failed: %v", tzErr)
 			// Fall back to marking completed with command results only
 			c.updateInvestigationStatus(investigation.ID, "completed", resultsForDB, nil)
 			return
 		}
 
 		if len(tzResp.Choices) == 0 {
-			fmt.Printf("⚠️  No choices in TensorZero response\n")
+			logging.Warning("No choices in TensorZero response")
 			c.updateInvestigationStatus(investigation.ID, "completed", resultsForDB, nil)
 			return
 		}
@@ -672,7 +674,7 @@ func (c *WebSocketClient) handlePendingInvestigation(investigation PendingInvest
 		if len(aiContent) > 300 {
 			// AI response received successfully
 		} else {
-			fmt.Printf("🤖 AI Response: %s\n", aiContent)
+			logging.Debug("AI Response: %s", aiContent)
 		}
 
 		// Check if this is a resolution response (final)
@@ -683,14 +685,14 @@ func (c *WebSocketClient) handlePendingInvestigation(investigation PendingInvest
 			Confidence     string `json:"confidence"`
 		}
 
-		fmt.Printf("🔍 Analyzing AI response type...\n")
+		logging.Debug("Analyzing AI response type...")
 
 		if err := json.Unmarshal([]byte(aiContent), &resolutionResp); err == nil && resolutionResp.ResponseType == "resolution" {
 			// This is the final resolution - show summary and complete
-			fmt.Printf("\n=== DIAGNOSIS COMPLETE ===\n")
-			fmt.Printf("Root Cause: %s\n", resolutionResp.RootCause)
-			fmt.Printf("Resolution Plan: %s\n", resolutionResp.ResolutionPlan)
-			fmt.Printf("Confidence: %s\n", resolutionResp.Confidence)
+			logging.Info("=== DIAGNOSIS COMPLETE ===")
+			logging.Info("Root Cause: %s", resolutionResp.RootCause)
+			logging.Info("Resolution Plan: %s", resolutionResp.ResolutionPlan)
+			logging.Info("Confidence: %s", resolutionResp.Confidence)
 			finalAIContent = aiContent
 			break
 		}
@@ -703,7 +705,7 @@ func (c *WebSocketClient) handlePendingInvestigation(investigation PendingInvest
 		}
 
 		if err := json.Unmarshal([]byte(aiContent), &diagnosticResp); err == nil && diagnosticResp.ResponseType == "diagnostic" {
-			fmt.Printf("🔄 AI requested additional diagnostics, executing...\n")
+			logging.Debug("AI requested additional diagnostics, executing...")
 
 			// Execute additional commands if any
 			additionalResults := map[string]interface{}{
@@ -711,7 +713,7 @@ func (c *WebSocketClient) handlePendingInvestigation(investigation PendingInvest
 			}
 
 			if len(diagnosticResp.Commands) > 0 {
-				fmt.Printf("🔧 Executing %d additional diagnostic commands...\n", len(diagnosticResp.Commands))
+				logging.Debug("Executing %d additional diagnostic commands", len(diagnosticResp.Commands))
 				commandResults := c.executeCommandsFromPayload(diagnosticResp.Commands)
 				additionalResults["command_results"] = commandResults
 			}
@@ -738,7 +740,7 @@ func (c *WebSocketClient) handlePendingInvestigation(investigation PendingInvest
 		}
 
 		// If neither resolution nor diagnostic, treat as final response
-		fmt.Printf("⚠️  Unknown response type - treating as final response\n")
+		logging.Warning("Unknown response type - treating as final response")
 		finalAIContent = aiContent
 		break
 	}
@@ -814,21 +816,21 @@ func (c *WebSocketClient) attemptReconnection() {
 
 			// Only show messages after 5 consecutive failures
 			if c.consecutiveFailures >= 5 {
-				log.Printf("🔄 Attempting WebSocket reconnection (attempt %d/%d) - %d consecutive failures", i+1, len(backoffDurations), c.consecutiveFailures)
+				logging.Info("Attempting WebSocket reconnection (attempt %d/%d) - %d consecutive failures", i+1, len(backoffDurations), c.consecutiveFailures)
 			}
 
 			time.Sleep(backoff)
 
 			if err := c.connect(); err != nil {
 				if c.consecutiveFailures >= 5 {
-					log.Printf("❌ Reconnection attempt %d failed: %v", i+1, err)
+					logging.Warning("Reconnection attempt %d failed: %v", i+1, err)
 				}
 				continue
 			}
 
 			// Successfully reconnected - reset failure counter
 			if c.consecutiveFailures >= 5 {
-				log.Printf("✅ WebSocket reconnected successfully after %d failures", c.consecutiveFailures)
+				logging.Info("WebSocket reconnected successfully after %d failures", c.consecutiveFailures)
 			}
 			c.consecutiveFailures = 0
 			go c.handleMessages() // Restart message handling
@@ -836,5 +838,5 @@ func (c *WebSocketClient) attemptReconnection() {
 		}
 	}
 
-	log.Printf("❌ Failed to reconnect after %d attempts, giving up", len(backoffDurations))
+	logging.Error("Failed to reconnect after %d attempts, giving up", len(backoffDurations))
 }