Show HN: Orchestrate parallel Claude Code and Codex agents on a live map (github.com)

A new tool called "agent-log-viewer" has been released, enabling users to orchestrate and visualize interactions between Codex and Claude Code agents through a local web interface. This application transforms raw logs into a user-friendly, live-updating chat feed, displaying each session and its subagents as a structured tree while providing real-time insights into ongoing processes. The tool operates entirely locally, reading existing log files without requiring a database or external services. This development is significant for the AI/ML community as it enhances the monitoring and management of AI agent sessions, facilitating improved collaboration and debugging workflows. Key features include support for multi-language UIs, deep linking for specific logs, and a project management layout that visually connects parent conversations to spawned agents. By leveraging direct access to Codex and Claude Code logs, as well as enhanced session orchestration capabilities, the agent-log-viewer promises to ease the implement-and-review cycles, ultimately optimizing the integration of AI models into various applications.