Long-term memory for AI agents via commonplace CLI — store, search, and recall facts across sessions
Long-term memory via commonplace CLI. Use when you learn something worth remembering (preferences, decisions, errors, context), at the start of tasks to recall past context, or when the user asks you to remember or recall something.
Install commonplace hooks into ~/.claude/settings.json. Run once after installing the plugin. Copies session-start and session-summarize hooks, then wires them into Stop and SessionStart events.
Every time you start a Claude Code session, it doesn't know what you decided last week. This plugin fixes that.
It's a Claude Code plugin for commonplace. It teaches Claude when and how to use the commonplace CLI, and wires in two session hooks: one that pulls relevant memories when a session starts, and one that saves what happened when it ends.
Install the commonplace binary first:
curl -fsSL https://raw.githubusercontent.com/saikatkumardey/commonplace/master/install.sh | sh
claude plugin add saikatkumardey/commonplace-plugin
Then run the setup skill once to install the hooks:
/commonplace setup
After setup, Claude Code will automatically:
session-log (Stop hook)preferences, decisions, errors, and context topics| Skill | Trigger | What it does |
|---|---|---|
/commonplace | Memory tasks | Teaches Claude the CLI interface and when to use it |
/commonplace setup | Run once after install | Copies hooks to ~/.claude/hooks/ and wires them into settings.json |
The setup skill installs two hooks.
SessionStart — hooks/session-start.py
Injects recent commonplace entries as additional context at the start of every session. Claude sees your past decisions, preferences, and context before responding to the first message.
Stop — hooks/session-summarize.py
When the session ends, reads the session transcript and uses claude -p (Haiku) to:
session-logpreferences, decisions, errors, and context topics — only entries genuinely worth remembering, with keyword-rich text for BM25 searchRuns async so it doesn't block exit.
commonplace write <topic> <entry> # save a memory
commonplace read <topic> # read a full topic
commonplace search <query> # BM25 search across all topics
commonplace topics # list all topics
commonplace forget <topic> <search> # remove entries
Memories are stored as plain markdown files in ~/.commonplace/. Human-readable and human-editable.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
npx claudepluginhub saikatkumardey/commonplace-plugin --plugin commonplaceThe bridge between Claude's working memory and Basic Memory's durable knowledge graph — session briefings, pre-compaction checkpoints, and capture reflexes
Long-term memory for Claude Code — slash commands, skills, opt-in session hooks, codebase indexing, and a deep-retrieval subagent over the @memsy-io/mcp server.
Cross-host durable memory — same ling-mem daemon and store in Claude Code, Codex, OpenClaw, and Linggen. Three-tier model (core + long-term + episodic staging) of who the user is, not a log of what was done.
Continuous memory for Claude Code. Extracts, summarizes, and compresses conversations into tiered daily logs. Claude remembers what you did yesterday.
Long-term semantic memory for Claude Code, powered by OpenViking. Auto-recall relevant memories at session start and capture important information during conversations.
Ultra-compressed communication mode. Cuts ~75% of tokens while keeping full technical accuracy by speaking like a caveman.