$ loading_
Manage branchable LLM history and encrypted context capsules for agents.
Copy the install command and let the AI configure it · recommended for beginners
Please install the "io.github.Medhovarsh/forkmind" MCP server from askskill: Run: claude mcp add 'io-github-medhovarsh-forkmind' -- npx -y forkmind
Developers or researchers can branch the same history into multiple versions when testing prompts, tool calls, or decision paths. This helps compare outcomes without overwriting the original session.
When an agent needs to pause and continue later, it can save context as an encrypted capsule and restore it afterward. This helps preserve session state while keeping context manageable.
In agent workflows involving confidential material, saved context can be crypto-shredded. This is useful for reducing sensitive data retention after a task is finished.
It provides a branchable LLM history DAG plus encrypted context capsules that agents can save, restore, and crypto-shred.
Based on the description, it supports saving context, restoring context, and crypto-shredding context. For any additional operations, see the source repository.
The provided material does not specify installation steps, runtime requirements, or key requirements. Please see the source repository for prerequisites.
Give all your AI agents shared context, memory, and coordinated reasoning.
Combine context-engineering methods to help AI agents gather and organize context.
Record and restore agent context with fan-out packs and signed receipts.
Share, search, and reuse knowledge across connected AI agents.
Encrypted portable memory for AI agents with cross-machine sync.
Store AI context locally and resume work seamlessly across tools.