Progress: Attended the Hack Your Summer project review (there are great projects this cohort! 🎉) and Vivek's talk (inspiration). Experimented with gathering USGS raw water quality data via R and ended up building a custom Claude Code/AI Agent Skill using curl to query and summarize the USGS API, releasing a early version for feedback (Github).
Struggle: Reached a potential strategic crossroads. While the AI agent makes complex USGS source-water querying significantly easier, it covers the source-to-treatment stage rather than the treatment-to-tap municipal level targeted by the PDF extraction pipeline. Weighing whether to stick with the deterministic data pipeline or pivot toward the AI agent route.
Plan: Gather community feedback on the newly released AI agent skill. Evaluate which direction—the AI agent or the municipal data pipeline—best tells the project's story before fully committing the build focus.