<p>Every Claude session starts from zero. You re-explain the architecture, re-paste the plan, re-derive the decisions you already made last week. Context windows help inside a session, but they vanish the moment you close the tab.</p>
<p>Logseq Brain is a Claude Code plugin that fixes this by turning a <a href="https://logseq.com" target="_blank" rel="noopener nofollow">https://logseq.com</a> graph into Claude’s external brain. Project pages, decision logs, and session journals live as plain markdown in a graph</p>
<p>you own — and because Logseq syncs across devices, your context follows you from desktop to laptop to phone.</p>
<h2>How it works</h2>
<p>The plugin is just four markdown skills — no runtime code, no build step. Claude itself is the runtime, using its standard Read/Write/Edit tools to operate on the graph:</p>
<ul>
<li>brain-init — scaffold the graph, add new projects</li>
<li>brain-load — pull project context into the current session (brief, full, or cross-graph search)</li>
<li>brain-save — surgically append session logs, decisions, and plan updates</li>
<li>brain-status — dashboard across all tracked projects, flags stale ones</li>
</ul>
<h2>Why Logseq?</h2>
<ul>
<li>You own your data. Plain markdown files in a folder. No proprietary database, no vendor lock-in.</li>
<li>Cross-device sync for free. Logseq Sync handles the hard part.</li>
<li>Queryable structure. Namespaced pages, properties (status:: accepted), and backlinks make decisions traceable over time.</li>
<li>Human-readable. Open the graph in Logseq to browse your AI-collaboration history like a research notebook.</li>
</ul>
<h2>Install</h2>
<pre><code>claude plugin marketplace <span class="hljs-keyword">add </span><span class="hljs-keyword">jame581/skillsmith
</span>claude plugin <span class="hljs-keyword">install </span>logseq-<span class="hljs-keyword">brain@skillsmith
</span></code></pre>
<p>Also works with GitHub Copilot CLI and Gemini CLI.</p>
<p>Repo: <a href="https://github.com/jame581/LogseqBrain" target="_blank" rel="noopener nofollow">https://github.com/jame581/LogseqBrain</a></p>


JaMe

Logseq Brain is a Claude Code plugin that gives Claude persistent memory across sessions by using a Logseq graph as an external knowledge store. It consists of four markdown-based skills (brain-init, brain-load, brain-save, brain-status) that let Claude read and write project context, decision logs, and session journals stored as plain markdown files you own. Since Logseq syncs across devices, your AI collaboration context follows you everywhere. It also works with GitHub Copilot CLI and Gemini CLI.

Give Claude a Memory: Persistent Context for Claude Code via Logseq