Save high-value knowledge before a context reset via /wipe.
npx skills add https://github.com/Bumblebiber/hmem --skill hmem-wipe
Please help me install this Agent Skill. Skill: hmem-wipe Source: https://github.com/Bumblebiber/hmem/tree/main/skills/hmem-wipe Command: npx skills add https://github.com/Bumblebiber/hmem --skill hmem-wipe
Wipe provides a safe, structured way to prepare for a context reset by optionally saving high-value knowledge so critical insights aren't lost.
Type /wipe to prepare for the clear, then type /clear to automatically restore your project context.
High-intent search queries and answers about installing and using this skill.
You can save AI memory context before clearing by using a pre-clear checkpoint process that optionally saves high-value knowledge like lessons and decisions before resetting the session.
Context clearing in AI memory management resets the session to remove accumulated noise and prevent threshold overloads. You should use it when a context-threshold warning appears or when the conversation becomes too long to process effectively.
To restore project context automatically after wiping, run a clear command that triggers a SessionStart hook. This hook automatically reloads the project context and briefing without requiring manual re-entry of previous information.
Yes, the context clear workflow supports automatic checkpointing of decisions and errors according to configured checkpointMode settings. This ensures critical insights are safely captured before the reset operation is executed.
The best way to prevent memory loss during a context reset is to follow a guided two-step wipe workflow that saves high-value knowledge first, then clears the context and automatically restores the project environment upon restart.
zunyiqingfeng-codeSummarize chat content and save context for auto-restore via SessionStart hook.
shiimanCreate plugin hooks for PreToolUse, PostToolUse, and session events.
nihaoshiManage writing sessions by saving, restoring, and updating context.
BEKO2210Parse markdown files to manage an AI's multi-layered memory system.
