What agent state should you persist across sessions? Devs weigh in
ombre-ai · reddit · 2026-10-25
A Reddit discussion asks: for long-running agent work, what should actually survive when one session ends and the next begins?
The author argues against persisting full transcripts, listing candidate state instead: decisions made, constraints, rejected approaches, current state, next steps, and files/artifacts. The open question is where that state should live — files, database, memory store, or summaries.
The post solicits production practices from developers running projects that span multiple sessions; the value lies in the comment-thread exchange of real-world approaches.
More from coding & agent
- 10 open-source GitHub repos that put every AI model behind one API — victor_explore · 2026-10-25
- Viajante: Open-Source MCP Server Scrapes Real Flight and Hotel Prices, No API Keys — Fast_Entrance7839 · 2026-10-25
- HCI researcher bets AI can finally make Bayesian stats usable for small-N studies — IanArawjo · 2026-10-25
- iOS demo combines GPT-Live-1, Decisions API and Stream Video for real-time vision — amos_gyamfi · 2026-10-25
- Harness design drives AI inference energy: cache misses can double a request's energy use — bibryam · 2026-10-25
- Redditor builds R-rated comedy club where AI agents do stand-up — and yours can sign up — bshane54 · 2026-10-25