netra-zen 1.0.9__tar.gz → 1.0.11__tar.gz

{netra_zen-1.0.9 → netra_zen-1.0.11}/LICENSE.md

@@ -1 +1 @@
-© Netra Inc. All rights reserved. Use is subject to Netra's Terms of Service.
+© Netra Inc. All rights reserved. Use is subject to Netra's Terms of Service.

{netra_zen-1.0.9 → netra_zen-1.0.11}/MANIFEST.in

@@ -1,37 +1,37 @@
-# Include documentation
-include README.md
-include LICENSE
-include CHANGELOG.md
-include PACKAGING.md
-
-# Include configuration examples
-include *.json
-
-# Include requirements files
-include requirements*.txt
-
-# Include test files
-recursive-include tests *.py
-recursive-include tests *.json
-recursive-include tests *.yaml
-
-# Include documentation
-recursive-include docs *.md
-recursive-include docs *.rst
-
-# Include package data
-recursive-include agent_interface *.py
-recursive-include token_budget *.py
-recursive-include token_transparency *.py
-
-# Exclude unnecessary files
-global-exclude *.pyc
-global-exclude *.pyo
-global-exclude *.swp
-global-exclude .DS_Store
-global-exclude __pycache__
-exclude .gitignore
-exclude .github
-exclude tox.ini
-exclude .coverage
+# Include documentation
+include README.md
+include LICENSE
+include CHANGELOG.md
+include PACKAGING.md
+
+# Include configuration examples
+include *.json
+
+# Include requirements files
+include requirements*.txt
+
+# Include test files
+recursive-include tests *.py
+recursive-include tests *.json
+recursive-include tests *.yaml
+
+# Include documentation
+recursive-include docs *.md
+recursive-include docs *.rst
+
+# Include package data
+recursive-include agent_interface *.py
+recursive-include token_budget *.py
+recursive-include token_transparency *.py
+
+# Exclude unnecessary files
+global-exclude *.pyc
+global-exclude *.pyo
+global-exclude *.swp
+global-exclude .DS_Store
+global-exclude __pycache__
+exclude .gitignore
+exclude .github
+exclude tox.ini
+exclude .coverage
 exclude .pytest_cache

{netra_zen-1.0.9/netra_zen.egg-info → netra_zen-1.0.11}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: netra-zen
-Version: 1.0.9
+Version: 1.0.11
 Summary: Multi-instance Claude orchestrator for parallel task execution
 Home-page: https://github.com/netra-systems/zen
 Author:  Systems
@@ -60,39 +60,38 @@ It works by analyzing your usage logs for metadata optimizations. It is focused
 
 This is a micro startup effort, aiming to provide real value for individual devs in exchange for feedback. Our intent is to charge businesses for larger scale optimizations.
 
-The process is simple. One time install, then one command. It auto grabs the last 3 log files and provides actionable items to update going forward to get the value of the optimizations.
+The process is simple. One time install, then one command. It analyzes your most recent log file and provides actionable items to update going forward to get the value of the optimizations. For best results, analyze one log file at a time with payloads under 1MB for focused, accurate analysis.
 
 ## Quick start
 
 1. `pip install netra-zen`
-2. `zen --apex --send-logs`
+2. `zen --apex`  # Logs are now sent by default
 3. Read the results and update claude settings, prompts, commands, etc. as needed to benefit
 
 See detailed install below if needed.
 
 ### Log Collection Options
 
-The optimizer analyzes your Claude Code usage logs to identify optimization opportunities. You can customize what logs are sent:
+The optimizer automatically analyzes your Claude Code usage logs to identify optimization opportunities. You can customize the behavior:
 
 ```bash
-# Send logs from the 3 most recent files (default)
-zen --apex --send-logs
-
-# Send logs from more files for deeper analysis
-zen --apex --send-logs --logs-count 5
+# Send logs from the most recent file (default behavior)
+zen --apex
 
 # Send logs from a specific project
-zen --apex --send-logs --logs-project "my-project-name"
+zen --apex --logs-project "my-project-name"
 
 # Send logs from a custom location
-zen --apex --send-logs --logs-path "/path/to/.claude/Projects"
+zen --apex --logs-path "/path/to/.claude/Projects"
 
-# Combine options for targeted analysis
-zen --apex --send-logs --logs-count 3 --logs-project "production-app"
+# Advanced: Send multiple files (use with caution)
+zen --apex --logs-count 2
 ```
 
 **Important:**
-- `--logs-count` specifies the number of **files** to read, not entries
+- `--logs-count` default is **1** for current Alpha release limitations. 
+- For optimal analysis, use **1 log file at a time** and keep payload under 1MB
+- Multiple log files can dilute the analysis and reduce accuracy
 - Each file may contain many log entries
 - The tool will display exactly how many entries from how many files are being sent
 
@@ -110,7 +109,7 @@ This was just changing a few small lines on a 400 line command.
 
 ## Example output from single file
 ```
-zen --apex --send-logs --logs-path /Users/user/.claude/projects/-Users-Desktop-netra-apex/7ac6d7ac-abc3-4903-a482-......-1.jsonl
+zen --apex --logs-path /Users/user/.claude/projects/-Users-Desktop-netra-apex/7ac6d7ac-abc3-4903-a482-......-1.jsonl
 
 SUCCESS: WebSocket connected successfully!
 
@@ -237,6 +236,28 @@ Receiving events...
 📊 Received 8 events
 ```
 
+# Another example
+```
+"issue": "Confused execution path and backtracking",                                               │
+
+"evidence": "The model initially assumes a JavaScript/TypeScript frontend based on the term 
+│ \"client side\", leading to incorrect file searches. After realizing it's a Python project, it gets stuck  
+│ repeatedly grepping `scripts/agent_cli.py` and requires multiple user interventions to be redirected to    
+│ the actual root cause in `zen_orchestrator.py` and `zen/telemetry/apex_telemetry.py`",         ...                                                                          │
+│         "token_waste": "8000-10000",                                                                       │
+│         "fix": "When an initial search term (e.g., 'ultrathink') fails, the model should prioritize        │
+│ understanding the project's overall structure and language (`ls -R` or `find . -type f | head -n 20`)      │
+│ before making assumptions. For regressions, `git log` should be used earlier in the process to identify    │
+│ recent, relevant changes that could have introduced the bug.",        
+
+│         "ideal prompt": "There's a regression where events are batched instead of streamed. The user       │
+│ mentioned 'client side ultrathink', but that could be a red herring. Start by checking the project         │
+│ structure and language, then review the git history for recent changes related to event handling,          │
+│ streaming, telemetry, or subprocess execution before deep-diving into code.",                              │
+│         "priority": "high"                                                                                 │
+│       }
+```                                    
+
 # Advanced features & detailed install guide
 
 In addition to optimizing your costs and latency,