LLM applicationsActively maintained
Loading this repository…
Persistent Context Across Sessions for Every Agent – Captures everything your agent does during sessions, compresses it with AI, and injects relevant context back into future sessions. Works with Claude Code, OpenClaw, Codex, Gemini, Hermes, Copilot, OpenCode + MoreDescription from GitHub
This tool captures what an AI coding agent did in one session, compresses it, and re-injects the relevant parts into a future session — an attempt at giving an agent memory across otherwise-separate conversations.
It suits a heavy agent user who is tired of re-explaining project context every session and wants some of that context to persist automatically.
Persisted memory is also a new place for stale or wrong context to quietly influence future sessions, and reviewing what got stored takes its own discipline. It is a productivity layer on top of an agent, not a replacement for the agent itself.
| Month | Days appeared |
|---|---|
| Dec 25 | 7 |
| Jan 26 | 2 |
| Feb 26 | 5 |
| Mar 26 | 2 |
| Apr 26 | 5 |
| May 26 | 1 |
Appeared on a GitHub trending board 22 days between 1 Dec 2025 and 31 May 2026. This counts trending-board appearances — it reflects attention, not quality, and GitHub does not publish how trending is calculated.
Repository details from the GitHub API, collected 16 Aug 2026. Trending history from the GitHub Trending Archive, covering 1 Dec 2025 – 31 May 2026.
All AI repos