tad-hq

16 ms·

1. 1.▲
   by tad-hq 10mo ago

   If you look at the jsonl structure of the sessions inside of your .claude/projects directory you should be able to find the token usage you are looking for. It's saved directly at the end of every tool call.
2. 2.▲
   by tad-hq 10mo ago

   I went through the same stages—started with "dump everything," then "search it," and recently landed on "let the agent read it for me." I started this as a private project like 2 months ago very simply, and it
3. 3.▲
   by tad-hq 10mo ago

   That's a clever approach for centralization. The hook method is solid for logging. The key difference is you're doing full-text search on raw conversations. With my MCP approach, Claude gets both the raw history and AI-generated s
4. 4.▲
   Show HN: UI and MCP server for analyzing Claude Code history. No more lost ideas(github.com)

   3 points by tad-hq 10mo ago|7 comments