Workshop shows how to analyze Claude Code and Copilot traces in DuckDB
Al_Grigor · x · 2026-07-28
A workshop shows how coding-agent traces can be treated as data: local session metadata from tools like Claude Code, Codex, and Copilot can be ingested, queried, and visualized.
The workflow is straightforward:
- read local JSON/JSONL trace files
- load them with dlt
- store them in DuckDB
- analyze usage by day, model, and project
The talk also highlights practical questions these traces can answer, such as token consumption, model mix, project activity, and whether secrets appear in logs.
More from coding & agent
- NVIDIA Launches DeepStream 9.1 for Multi-Camera 3D Tracking Apps — NVIDIAAI · 2026-07-30
- Replit Agent Deleted Production DB Ignoring ALL CAPS: Why Prompts Aren't Guardrails — jayesh_ahire1 · 2026-07-30
- Small-model orchestration roughly doubled task completion in a 100-task benchmark — _raydeStar · 2026-07-29
- Video Talk: Safeguarding RAG & AI Agent Systems Using OpenFGA — HowDevelop · 2026-07-29
- Nous Portal bundles model access and a tool gateway for agent workflows — Teknium · 2026-07-29
- Reddit test says direct MCP beats Zapier for scheduling Claude social posts — Purple_Network3016 · 2026-07-29