netra-zen 1.0.10__py3-none-any.whl → 1.1.2__py3-none-any.whl

scripts/verify_log_transmission.py

@@ -1,140 +1,140 @@
-#!/usr/bin/env python3
-"""
-Verification script to prove JSONL logs are bundled in payload
-"""
-
-import json
-import sys
-from pathlib import Path
-
-# Add parent directory to path
-sys.path.insert(0, str(Path(__file__).parent.parent))
-
-from scripts.agent_logs import collect_recent_logs
-
-
-def verify_log_bundling(log_path: str):
-    """
-    Verify that logs are properly collected and bundled
-
-    Args:
-        log_path: Path to JSONL file or directory
-    """
-    print("=" * 70)
-    print("JSONL LOG TRANSMISSION VERIFICATION")
-    print("=" * 70)
-    print()
-
-    # Step 1: Collect logs
-    print("Step 1: Collecting logs from file...")
-    result = collect_recent_logs(limit=1, base_path=log_path)
-
-    if not result:
-        print("❌ FAILED: No logs collected")
-        return False
-
-    logs, files_read, file_info = result
-    print(f"✓ Successfully collected {len(logs)} log entries from {files_read} file(s)")
-    print()
-
-    # Step 2: Show file details
-    print("Step 2: File details...")
-    for info in file_info:
-        print(f"  File: {info['name']}")
-        print(f"  Hash: {info['hash']}")
-        print(f"  Entries: {info['entries']}")
-    print()
-
-    # Step 3: Simulate payload creation
-    print("Step 3: Simulating WebSocket payload creation...")
-    payload = {
-        "type": "message_create",
-        "run_id": "test-run-id",
-        "payload": {
-            "message": "Test message with logs",
-            "jsonl_logs": logs  # This is where logs are added
-        }
-    }
-
-    print(f"✓ Payload created with 'jsonl_logs' key")
-    print(f"  Payload keys: {list(payload['payload'].keys())}")
-    print()
-
-    # Step 4: Verify payload size
-    print("Step 4: Calculating payload size...")
-    payload_json = json.dumps(payload)
-    payload_size_bytes = len(payload_json.encode('utf-8'))
-    payload_size_kb = payload_size_bytes / 1024
-    payload_size_mb = payload_size_kb / 1024
-
-    if payload_size_mb >= 1:
-        size_str = f"{payload_size_mb:.2f} MB"
-    elif payload_size_kb >= 1:
-        size_str = f"{payload_size_kb:.2f} KB"
-    else:
-        size_str = f"{payload_size_bytes} bytes"
-
-    print(f"✓ Total payload size: {size_str}")
-    print()
-
-    # Step 5: Show sample log entries
-    print("Step 5: Sample log entries in payload...")
-    if logs:
-        print(f"  First entry keys: {list(logs[0].keys())}")
-        print(f"  First entry timestamp: {logs[0].get('timestamp', 'N/A')}")
-        print(f"  Last entry timestamp: {logs[-1].get('timestamp', 'N/A')}")
-    print()
-
-    # Step 6: Verify transmission-ready
-    print("Step 6: Transmission verification...")
-    print(f"✓ Payload is valid JSON: {payload_json is not None}")
-    print(f"✓ Payload contains 'jsonl_logs': {'jsonl_logs' in payload['payload']}")
-    print(f"✓ Log count in payload: {len(payload['payload']['jsonl_logs'])}")
-    print()
-
-    print("=" * 70)
-    print("✅ VERIFICATION COMPLETE")
-    print("=" * 70)
-    print()
-    print("PROOF OF TRANSMISSION:")
-    print(f"  • {len(logs)} JSONL log entries are bundled in the payload")
-    print(f"  • Payload size: {size_str}")
-    print(f"  • Ready for WebSocket transmission to backend")
-    print()
-
-    # Optional: Save proof file
-    proof_file = Path("/tmp/zen_transmission_proof.json")
-    proof_payload = {
-        "verification_timestamp": "verification_run",
-        "log_count": len(logs),
-        "files_read": files_read,
-        "file_info": file_info,
-        "payload_size": size_str,
-        "sample_first_entry": logs[0] if logs else None,
-        "sample_last_entry": logs[-1] if logs else None,
-        "payload_structure": {
-            "type": payload["type"],
-            "run_id": payload["run_id"],
-            "payload_keys": list(payload["payload"].keys()),
-            "jsonl_logs_present": "jsonl_logs" in payload["payload"],
-            "jsonl_logs_count": len(payload["payload"]["jsonl_logs"])
-        }
-    }
-
-    with open(proof_file, 'w') as f:
-        json.dump(proof_payload, f, indent=2)
-
-    print(f"📝 Detailed proof saved to: {proof_file}")
-    print()
-
-    return True
-
-
-if __name__ == "__main__":
-    if len(sys.argv) < 2:
-        print("Usage: python verify_log_transmission.py <path-to-jsonl-file>")
-        sys.exit(1)
-
-    log_path = sys.argv[1]
-    success = verify_log_bundling(log_path)
-    sys.exit(0 if success else 1)
+#!/usr/bin/env python3
+"""
+Verification script to prove JSONL logs are bundled in payload
+"""
+
+import json
+import sys
+from pathlib import Path
+
+# Add parent directory to path
+sys.path.insert(0, str(Path(__file__).parent.parent))
+
+from scripts.agent_logs import collect_recent_logs
+
+
+def verify_log_bundling(log_path: str):
+    """
+    Verify that logs are properly collected and bundled
+
+    Args:
+        log_path: Path to JSONL file or directory
+    """
+    print("=" * 70)
+    print("JSONL LOG TRANSMISSION VERIFICATION")
+    print("=" * 70)
+    print()
+
+    # Step 1: Collect logs
+    print("Step 1: Collecting logs from file...")
+    result = collect_recent_logs(limit=1, base_path=log_path)
+
+    if not result:
+        print("❌ FAILED: No logs collected")
+        return False
+
+    logs, files_read, file_info = result
+    print(f"✓ Successfully collected {len(logs)} log entries from {files_read} file(s)")
+    print()
+
+    # Step 2: Show file details
+    print("Step 2: File details...")
+    for info in file_info:
+        print(f"  File: {info['name']}")
+        print(f"  Hash: {info['hash']}")
+        print(f"  Entries: {info['entries']}")
+    print()
+
+    # Step 3: Simulate payload creation
+    print("Step 3: Simulating WebSocket payload creation...")
+    payload = {
+        "type": "message_create",
+        "run_id": "test-run-id",
+        "payload": {
+            "message": "Test message with logs",
+            "jsonl_logs": logs  # This is where logs are added
+        }
+    }
+
+    print(f"✓ Payload created with 'jsonl_logs' key")
+    print(f"  Payload keys: {list(payload['payload'].keys())}")
+    print()
+
+    # Step 4: Verify payload size
+    print("Step 4: Calculating payload size...")
+    payload_json = json.dumps(payload)
+    payload_size_bytes = len(payload_json.encode('utf-8'))
+    payload_size_kb = payload_size_bytes / 1024
+    payload_size_mb = payload_size_kb / 1024
+
+    if payload_size_mb >= 1:
+        size_str = f"{payload_size_mb:.2f} MB"
+    elif payload_size_kb >= 1:
+        size_str = f"{payload_size_kb:.2f} KB"
+    else:
+        size_str = f"{payload_size_bytes} bytes"
+
+    print(f"✓ Total payload size: {size_str}")
+    print()
+
+    # Step 5: Show sample log entries
+    print("Step 5: Sample log entries in payload...")
+    if logs:
+        print(f"  First entry keys: {list(logs[0].keys())}")
+        print(f"  First entry timestamp: {logs[0].get('timestamp', 'N/A')}")
+        print(f"  Last entry timestamp: {logs[-1].get('timestamp', 'N/A')}")
+    print()
+
+    # Step 6: Verify transmission-ready
+    print("Step 6: Transmission verification...")
+    print(f"✓ Payload is valid JSON: {payload_json is not None}")
+    print(f"✓ Payload contains 'jsonl_logs': {'jsonl_logs' in payload['payload']}")
+    print(f"✓ Log count in payload: {len(payload['payload']['jsonl_logs'])}")
+    print()
+
+    print("=" * 70)
+    print("✅ VERIFICATION COMPLETE")
+    print("=" * 70)
+    print()
+    print("PROOF OF TRANSMISSION:")
+    print(f"  • {len(logs)} JSONL log entries are bundled in the payload")
+    print(f"  • Payload size: {size_str}")
+    print(f"  • Ready for WebSocket transmission to backend")
+    print()
+
+    # Optional: Save proof file
+    proof_file = Path("/tmp/zen_transmission_proof.json")
+    proof_payload = {
+        "verification_timestamp": "verification_run",
+        "log_count": len(logs),
+        "files_read": files_read,
+        "file_info": file_info,
+        "payload_size": size_str,
+        "sample_first_entry": logs[0] if logs else None,
+        "sample_last_entry": logs[-1] if logs else None,
+        "payload_structure": {
+            "type": payload["type"],
+            "run_id": payload["run_id"],
+            "payload_keys": list(payload["payload"].keys()),
+            "jsonl_logs_present": "jsonl_logs" in payload["payload"],
+            "jsonl_logs_count": len(payload["payload"]["jsonl_logs"])
+        }
+    }
+
+    with open(proof_file, 'w') as f:
+        json.dump(proof_payload, f, indent=2)
+
+    print(f"📝 Detailed proof saved to: {proof_file}")
+    print()
+
+    return True
+
+
+if __name__ == "__main__":
+    if len(sys.argv) < 2:
+        print("Usage: python verify_log_transmission.py <path-to-jsonl-file>")
+        sys.exit(1)
+
+    log_path = sys.argv[1]
+    success = verify_log_bundling(log_path)
+    sys.exit(0 if success else 1)

zen/__init__.py

@@ -1,7 +1,7 @@
-"""Zen namespace package placeholder.
-
-This lightweight module exists so repository scripts can import
-`zen.telemetry` directly without requiring the full orchestrator package.
-"""
-
-__all__: list[str] = []
+"""Zen namespace package placeholder.
+
+This lightweight module exists so repository scripts can import
+`zen.telemetry` directly without requiring the full orchestrator package.
+"""
+
+__all__: list[str] = []

zen/__main__.py

@@ -1,11 +1,11 @@
-"""Module entry point to support `python -m zen` invocation."""
-
-from zen_orchestrator import run
-
-
-def main() -> None:
-    run()
-
-
-if __name__ == "__main__":
-    main()
+"""Module entry point to support `python -m zen` invocation."""
+
+from zen_orchestrator import run
+
+
+def main() -> None:
+    run()
+
+
+if __name__ == "__main__":
+    main()

zen/telemetry/__init__.py

@@ -1,11 +1,14 @@
-"""Telemetry utilities exposed by the Zen package."""
-
-from .embedded_credentials import get_embedded_credentials, get_project_id
-from .manager import TelemetryManager, telemetry_manager
-
-__all__ = [
-    "TelemetryManager",
-    "telemetry_manager",
-    "get_embedded_credentials",
-    "get_project_id",
-]
+"""Telemetry utilities exposed by the Zen package."""
+
+from .embedded_credentials import get_embedded_credentials, get_project_id
+from .manager import TelemetryManager, telemetry_manager
+from .apex_telemetry import run_apex_with_telemetry, ApexTelemetryWrapper
+
+__all__ = [
+    "TelemetryManager",
+    "telemetry_manager",
+    "get_embedded_credentials",
+    "get_project_id",
+    "run_apex_with_telemetry",
+    "ApexTelemetryWrapper",
+]

zen/telemetry/apex_telemetry.py

@@ -0,0 +1,259 @@
+"""Telemetry wrapper for apex instance tracking.
+
+This module provides a lightweight wrapper around agent_cli.py subprocess calls
+to emit OpenTelemetry spans for apex instances without modifying agent_cli.py.
+"""
+
+from __future__ import annotations
+
+import json
+import logging
+import os
+import subprocess
+import sys
+import time
+from typing import Any, Dict, Optional
+
+from .manager import telemetry_manager
+
+logger = logging.getLogger(__name__)
+
+
+class ApexTelemetryWrapper:
+    """Wrapper to track apex instance telemetry."""
+
+    def __init__(self):
+        self.start_time: Optional[float] = None
+        self.end_time: Optional[float] = None
+        self.exit_code: Optional[int] = None
+        self.message: Optional[str] = None
+        self.env: str = "staging"
+        self.stdout: str = ""
+        self.stderr: str = ""
+
+    def run_apex_with_telemetry(
+        self,
+        agent_cli_path: str,
+        filtered_argv: list,
+        env: Optional[Dict[str, str]] = None
+    ) -> int:
+        """Run agent_cli.py subprocess and emit telemetry span.
+
+        Args:
+            agent_cli_path: Path to agent_cli.py script
+            filtered_argv: Command-line arguments (without 'zen' and '--apex')
+            env: Environment variables to pass to subprocess
+
+        Returns:
+            Exit code from agent_cli subprocess
+        """
+        self.start_time = time.time()
+
+        # Extract message from argv for telemetry
+        self.message = self._extract_message(filtered_argv)
+        self.env = self._extract_env(filtered_argv)
+
+        # Build command
+        cmd = [sys.executable, agent_cli_path] + filtered_argv
+
+        try:
+            # Use Popen for real-time streaming while still capturing output for telemetry
+            process = subprocess.Popen(
+                cmd,
+                env=env,
+                stdout=subprocess.PIPE,
+                stderr=subprocess.PIPE,
+                text=True,
+                bufsize=1  # Line buffered for real-time output
+            )
+
+            # Collect output while streaming in real-time
+            stdout_lines = []
+            stderr_lines = []
+
+            # Stream stdout in real-time
+            if process.stdout:
+                for line in iter(process.stdout.readline, ''):
+                    if line:
+                        print(line, end='')  # Print immediately for real-time display
+                        stdout_lines.append(line)
+
+            # Wait for process to complete and get stderr
+            stderr_output = process.stderr.read() if process.stderr else ""
+            if stderr_output:
+                print(stderr_output, end='', file=sys.stderr)
+                stderr_lines.append(stderr_output)
+
+            # Wait for process to complete
+            self.exit_code = process.wait()
+
+            # Store captured output for telemetry parsing
+            self.stdout = ''.join(stdout_lines)
+            self.stderr = ''.join(stderr_lines)
+
+        except Exception as e:
+            logger.warning(f"Failed to run apex subprocess: {e}")
+            self.exit_code = 1
+            self.stderr = str(e)
+
+        finally:
+            self.end_time = time.time()
+            self._emit_telemetry()
+
+        return self.exit_code or 0
+
+    def _extract_message(self, argv: list) -> str:
+        """Extract message from command-line arguments."""
+        try:
+            if '--message' in argv:
+                idx = argv.index('--message')
+                if idx + 1 < len(argv):
+                    return argv[idx + 1]
+            elif '-m' in argv:
+                idx = argv.index('-m')
+                if idx + 1 < len(argv):
+                    return argv[idx + 1]
+        except (ValueError, IndexError):
+            pass
+        return "apex-instance"
+
+    def _extract_env(self, argv: list) -> str:
+        """Extract environment from command-line arguments."""
+        try:
+            if '--env' in argv:
+                idx = argv.index('--env')
+                if idx + 1 < len(argv):
+                    return argv[idx + 1]
+        except (ValueError, IndexError):
+            pass
+        return "staging"
+
+    def _emit_telemetry(self) -> None:
+        """Emit OpenTelemetry span for apex instance."""
+        if telemetry_manager is None or not hasattr(telemetry_manager, "is_enabled"):
+            logger.debug("Telemetry manager not available")
+            return
+
+        if not telemetry_manager.is_enabled():
+            logger.debug("Telemetry is not enabled")
+            return
+
+        # Calculate duration
+        duration_ms = 0
+        if self.start_time and self.end_time:
+            duration_ms = int((self.end_time - self.start_time) * 1000)
+
+        # Determine status
+        status = "completed" if self.exit_code == 0 else "failed"
+        success = self.exit_code == 0
+
+        # Build attributes for apex.instance span
+        attributes: Dict[str, Any] = {
+            "zen.instance.type": "apex",
+            "zen.instance.name": "apex.instance",
+            "zen.instance.status": status,
+            "zen.instance.success": success,
+            "zen.instance.duration_ms": duration_ms,
+            "zen.instance.exit_code": self.exit_code or 0,
+            "zen.apex.environment": self.env,
+            "zen.apex.message": self._truncate_message(self.message or ""),
+        }
+
+        # Parse JSON output if available (contains token/cost info)
+        json_output = self._parse_json_output()
+        if json_output:
+            self._add_json_metrics(attributes, json_output)
+
+        # Emit span using the telemetry manager's tracer (same way as regular zen instances)
+        try:
+            # Access the tracer the same way telemetry_manager.record_instance_span() does
+            if not hasattr(telemetry_manager, '_tracer') or telemetry_manager._tracer is None:
+                logger.warning("Telemetry manager has no tracer configured")
+                return
+
+            from opentelemetry.trace import SpanKind
+            from google.api_core.exceptions import GoogleAPICallError
+
+            with telemetry_manager._tracer.start_as_current_span(
+                "apex.instance", kind=SpanKind.INTERNAL
+            ) as span:
+                for key, value in attributes.items():
+                    span.set_attribute(key, value)
+
+            logger.info(f"✅ Emitted apex telemetry span with {len(attributes)} attributes")
+            logger.debug(f"Apex span attributes: {attributes}")
+
+            # Note: Removed force_flush to prevent blocking event streaming
+            # Spans will still be sent via the normal batch export process
+
+        except Exception as exc:
+            logger.error(f"❌ Failed to emit apex telemetry span: {exc}")
+            import traceback
+            logger.debug(f"Traceback: {traceback.format_exc()}")
+
+    def _truncate_message(self, message: str, max_length: int = 200) -> str:
+        """Truncate message for telemetry attributes."""
+        if len(message) <= max_length:
+            return message
+        return message[:max_length] + "..."
+
+    def _parse_json_output(self) -> Optional[Dict[str, Any]]:
+        """Parse JSON output from agent_cli stdout if available."""
+        if not self.stdout:
+            return None
+
+        # Try to find JSON in stdout
+        for line in self.stdout.split('\n'):
+            line = line.strip()
+            if line.startswith('{') and line.endswith('}'):
+                try:
+                    return json.loads(line)
+                except json.JSONDecodeError:
+                    continue
+
+        return None
+
+    def _add_json_metrics(self, attributes: Dict[str, Any], json_output: Dict[str, Any]) -> None:
+        """Add metrics from JSON output to telemetry attributes."""
+        # Extract token usage if available
+        if 'usage' in json_output:
+            usage = json_output['usage']
+            attributes['zen.tokens.total'] = usage.get('total_tokens', 0)
+            attributes['zen.tokens.input'] = usage.get('input_tokens', 0)
+            attributes['zen.tokens.output'] = usage.get('output_tokens', 0)
+            attributes['zen.tokens.cache.read'] = usage.get('cache_read_tokens', 0)
+            attributes['zen.tokens.cache.creation'] = usage.get('cache_creation_tokens', 0)
+
+        # Extract cost if available
+        if 'cost' in json_output:
+            cost = json_output['cost']
+            if 'total_usd' in cost:
+                attributes['zen.cost.usd_total'] = round(float(cost['total_usd']), 6)
+
+        # Extract run_id if available
+        if 'run_id' in json_output:
+            attributes['zen.apex.run_id'] = json_output['run_id']
+
+        # Extract validation status
+        if 'validation' in json_output:
+            validation = json_output['validation']
+            attributes['zen.apex.validation.passed'] = validation.get('passed', False)
+
+
+def run_apex_with_telemetry(
+    agent_cli_path: str,
+    filtered_argv: list,
+    env: Optional[Dict[str, str]] = None
+) -> int:
+    """Convenience function to run apex with telemetry tracking.
+
+    Args:
+        agent_cli_path: Path to agent_cli.py script
+        filtered_argv: Command-line arguments (without 'zen' and '--apex')
+        env: Environment variables to pass to subprocess
+
+    Returns:
+        Exit code from agent_cli subprocess
+    """
+    wrapper = ApexTelemetryWrapper()
+    return wrapper.run_apex_with_telemetry(agent_cli_path, filtered_argv, env)