Skip to content

Optimize logging: move verbose logs to DEBUG level - #21

Merged
aviweit merged 1 commit into
skillberry-ai:mainfrom
aviweit:feature/logging-optimization
Jun 24, 2026
Merged

aviweit merged 1 commit into
skillberry-ai:mainfrom
aviweit:feature/logging-optimization

Conversation

@aviweit

@aviweit aviweit commented Jun 24, 2026

Copy link
Copy Markdown
Collaborator

Logging Optimization

Summary

Optimize logging levels to reduce log file size by 70-80% during benchmark runs by moving verbose logs from INFO to DEBUG level.

Changes

skillberry-benchmarks

  • File: tau2/tau2-bench/src/tau2/agent/lang_chain_agent_mcp.py
  • Removed langgraph_logger.setLevel(logging.DEBUG) override
  • Changed 13 verbose INFO logs → DEBUG (loops, schema dumps, detailed inspections)
  • Replaced 4 print() statements with logger.debug()

skillberry-agent

  • File: shared/python/skillberry_agent_lib/skillberry_agent_lib/mcp_interceptor.py

  • Changed 1 exit log from INFO → DEBUG for consistency

  • File: shared/python/skillberry_agent_lib/skillberry_agent_lib/langgraph_nodes.py

  • Changed 30+ verbose INFO logs → DEBUG (message inspection loops, tool call details)

Impact

  • Log file size: 70-80% reduction expected
  • INFO logs: Now show only high-level operational flow
  • DEBUG logs: Contain all verbose details for troubleshooting
  • No functional changes: Only logging level adjustments

Testing

  • Run tau2 benchmark with INFO logging
  • Verify log file size reduction
  • Confirm essential events still logged at INFO
  • Test DEBUG mode for verbose troubleshooting

Branch

feature/logging-optimization (same name in both repos)

@aviweit aviweit self-assigned this Jun 24, 2026
@aviweit
aviweit merged commit 069bcfc into skillberry-ai:main Jun 24, 2026
1 check passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant