Termem brings cross-agent memory sharing to the terminal
Termem is a terminal tool that enables quick agent session recovery and cross-agent memory sharing within a directory, accessible via MCP and a skill interface.
Score breakdown
Termem allows different AI coding agents to share session history within a directory, removing the isolation that normally prevents one agent from seeing another's prior context.
- 01Built by todience and posted to Hacker News on June 15, 2026.
- 02Running `termem` in a directory lists all agent sessions for that folder and allows resuming them.
- 03New agent sessions can recall all prior sessions in the same directory via MCP and a skill.
Termem is a terminal utility designed for cross-agent memory and session management. The core workflow involves navigating into a project directory and running `termem` to view all agent sessions associated with that folder, with the ability to resume any of them directly.
The tool also addresses cross-agent memory sharing: when a new agent session is launched, the agent can access all prior sessions within the same directory through an MCP integration and a skill interface.
The tool also addresses cross-agent memory sharing: when a new agent session is launched, the agent can access all prior sessions within the same directory through an MCP integration and a skill interface. This means an agent like Claude can read sessions that were originally created by Codex or other agents, enabling continuity across different AI coding assistants working in the same project context.
Key facts
- 01Built by todience and posted to Hacker News on June 15, 2026.
- 02Running `termem` in a directory lists all agent sessions for that folder and allows resuming them.
- 03New agent sessions can recall all prior sessions in the same directory via MCP and a skill.
- 04Cross-agent memory sharing is supported — Claude can view sessions created by Codex and other agents.
- 05The project is accessible at termem.com.
Topics
Summary and scoring are generated automatically from the original article. We always link back to the publisher and never republish images or paywalled content. Last processed Jun 16, 2026 · 23:11 UTC. How this works →