🧠 LogBrain

LogBrain is a smart Python log analysis toolkit that merges lightweight machine learning, symbolic reasoning, and story-driven flow graphs. It analyzes logs (text or screenshots), understands their structure, and generates meaningful summaries or problem statements using rule-based or mini-LLM methods.

🧩 Key Components

1. ConfigurablePatternType

• Parses log content using fuzzy logic or ML.
• Can match entries to pre-defined labels like CRON, APP, etc.
• Fully configurable via pattern_config.json.

2. CognitiveGraph

• Accepts a story (e.g. application flow) as a .txt file or LLM-generated text.
• Builds a flow map and uses it to detect the stage where errors occur.

📥 Sample Input: Logs

ERROR: cron job failed at 02:30
FAIL: could not execute script

🧠 Sample Summary (Generated via Tiny LLM)

The cron job could not execute the script at 02:30.
This failure likely caused the application timeout due to unmet preconditions.

📝 License

This project is licensed under the MIT License. See LICENSE for full details.

💻 Usage Example

from logbrain.logbrain import LogBrain

logbrain = LogBrain("pattern_config.json", "story.txt", is_file=True)
lines = ["ERROR: cron job failed at 02:30", "FAIL: could not execute script"]
logbrain.analyze_log("log.txt", lines)
logbrain.finalize(use_llm=True)  # Uses mini LLM to build better summary
print(logbrain.summarize())

📁 Files

• logbrain.py: Core logic
• pattern_config.json: Configurable pattern matching
• story.txt: Sample flow for CognitiveGraph
• test_logbrain.py: Test case
• setup.py, requirements.txt: Packaging

📦 Install & Run

pip install -r requirements.txt
python test_logbrain.py

Ready to log with brains! 🧠