Show HN: UsageGuru – Cut log noise and save context in Claude Code (usageguru.com)

UsageGuru has launched a new plugin designed to enhance the utility of Claude Code by significantly reducing log noise during debugging. By filtering out up to 90% of irrelevant information, such as repeated lines and framework spam, UsageGuru allows Claude to focus on critical errors without unnecessary clutter. This dramatically decreases the token usage, which is vital for users of Claude's AI, ultimately optimizing interactions and saving costs associated with token consumption. For the AI/ML community, this tool is significant as it directly addresses the challenges of handling large logs and stack traces, which can be cumbersome and lead to wasted computational resources. The plugin operates locally on users' machines, ensuring that sensitive code and data remain secure. It supports a range of AI tools that work with the MCP server model and can be integrated effortlessly with a simple command, making it accessible to developers looking to streamline their debugging processes. With a nominal fee of $5/month and the potential for cost savings, UsageGuru promises substantial value for daily users.