The latest articles on DEV Community by Ruca. (@ruca_ai). https://dev.to/ruca_ai https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4074982%2Ff1868a7e-3f00-4b25-b8bb-5793a460b71f.jpg https://dev.to/ruca_ai en Ruca. Wed, 12 Aug 2026 15:02:52 +0000 https://dev.to/ruca_ai/i-built-tracemotive-a-local-first-debugger-for-ai-agent-execution-29ek https://dev.to/ruca_ai/i-built-tracemotive-a-local-first-debugger-for-ai-agent-execution-29ek <p>I just released TraceMotive v0.1, an open-source, local-first tracing and debugging tool for AI agent execution.</p> <p>I built it because I kept running into the same problem:</p> <p>When an agent behaves incorrectly, the final error often doesn't tell you enough about how the execution actually got there.</p> <p>TraceMotive records agent runs as traces and spans so you can inspect the execution step by step.</p> <h2> What v0.1 includes </h2> <ul> <li>Trace and Span collection</li> <li>Parent/child Span hierarchy</li> <li>Execution Timeline</li> <li>Span Inspector</li> <li>SQLite-backed local Collector</li> <li>OpenAI Agents SDK integration</li> <li>Privacy-first content capture defaults</li> <li>Bounded local transport and failure isolation</li> </ul> <h2> Install </h2> <p>pip install tracemotive</p> <p>GitHub:<br> <a href="https://github.com/doraemonfv-glitch/tracemotive" rel="noopener noreferrer">https://github.com/doraemonfv-glitch/tracemotive</a></p> <p>PyPI:<br> <a href="https://pypi.org/project/tracemotive/" rel="noopener noreferrer">https://pypi.org/project/tracemotive/</a></p> <p>TraceMotive's own tracing data stays local by default. Model-provider traffic still depends on how the agent itself is configured.</p> <h2> Why I built it </h2> <p>v0.1 is intentionally focused on observation and debugging rather than automatic diagnosis.</p> <p>The longer-term direction is to move from:</p> <p>“Where did the error happen?”</p> <p>toward:</p> <p>“Where did the execution first start going wrong?”</p> <p>That is the idea behind the long-term goal:</p> <p>“The causal debugger for AI agents.”</p> <p>I'd especially appreciate feedback from people building agents:</p> <p>What information do you usually wish you had when debugging an agent execution?</p> showdev ai agents python