Zikkaron FAQs

Question 1

What are the MCP tools provided by Zikkaron?

Accepted Answer

Zikkaron exposes 18 specialized tools via the Model Context Protocol (MCP) to give Claude fine-grained control over its memory. These tools facilitate memory management (`remember`, `forget`), context retrieval (`recall`, `get_project_context`), metacognition (`assess_coverage`), prospective triggers, and more.

Question 2

Can Zikkaron understand cause-effect relationships in my code?

Accepted Answer

Yes, Zikkaron features causal discovery capabilities, utilizing algorithms like PC to learn cause-effect relationships from your coding sessions and event logs. This helps Claude understand the deeper 'why' behind decisions and changes.

Question 3

What is Zikkaron and how does it help Claude Code?

Accepted Answer

Zikkaron is a biologically-inspired persistent memory engine for Claude Code. It solves the problem of Claude forgetting past interactions by learning, consolidating, and surfacing relevant context across coding sessions, eliminating the need to re-explain project details.

Question 4

How does Zikkaron implement 'biologically-inspired' memory?

Accepted Answer

Zikkaron incorporates advanced concepts like predictive coding for smart ingest, heat-based salience for natural memory decay, reconsolidation for dynamic updates, fractal hierarchies for multi-scale summarization, and a knowledge graph for rich context retrieval, mimicking human cognitive processes.

Question 5

Is Zikkaron's memory storage local and private?

Accepted Answer

Yes, absolutely. All data, including embeddings and knowledge graphs, is stored entirely on your local machine in a single SQLite database using FTS5 and `sqlite-vec`. It functions offline with no cloud dependencies, API calls, or telemetry.

Zikkaron

Zikkaron

Características Principales

Casos de Uso

Características Principales

Casos de Uso