Motive-mem
After each turn, another model reviews the diff against your repo principles for intent, violations, and band-aids.
Illustration. Actual results depend on the mod, terminal and environment.
Install
/plugin marketplace add wjsekrud/motive-mem
/plugin install motive-mem@motive-mem
/reload-pluginsRun these commands one at a time in a Claude Code session.
claude plugin marketplace add wjsekrud/motive-mem
claude plugin install motive-mem@motive-memThen run /reload-plugins in your session, or start a new session.
Please install the Claude Code mod "Motive-mem" (https://github.com/wjsekrud/motive-mem). Before installing, read the Security and Requirements sections of its README and tell me what it reads and writes, which commands it runs, whether it uses the network, and which Claude Code version, OS and terminal it needs. If everything looks fine, run claude plugin marketplace add wjsekrud/motive-mem and claude plugin install motive-mem@motive-mem, then remind me to run /reload-plugins in my session.Send this to Claude Code so it reads the author’s instructions before installing.
git clone https://github.com/wjsekrud/motive-mem
cd motive-mem
claude --plugin-dir .Clone the repository and try it in a local session with --plugin-dir.
What it accesses
Based on the author’s documentation, checked when listed; use claude plugin validate to check for yourself.
Reads
YesReads the principles file, turn diffs, related source files, and approved exceptions.
Writes files
YesWrites review logs and exception decisions under .git/motive-mem/ (outside the work tree).
Runs commands
YesRuns git (including stash create) for snapshots and diffs without changing the work tree.
Network
YesCalls $.model.complete for a separate review after code-changing turns (model requests).
Introduction
Motive-mem is a Claude Code function-hooks mod: when a turn ends, a separate model reviews that turn's diff against .claude/principles.md for intent match, principle violations, and band-aid fixes, and shows findings in a panel.
- When it runs: only in a git repo with a principles file; comment/docs-only turns are skipped. Snapshots use
git stash createwithout touching the working tree. - Panel: send findings to Claude, review again, or approve as an exception (stored under
.git/motive-mem/decisions/). - Cost: one extra model call per code-changing turn (default sonnet); principles are prompt-cached.
- Commands:
/motive-mem,/motive-mem all,pause/resume; skill/motive-mem:principleshelps mine principles.
Marketplace wjsekrud/motive-mem. Author built/tested on Claude Code 2.1.299. Preview is illustrative.
Who it is for
- Teams with written principles who want each turn checked against them
Requirements
- Claude Code ≥ 2.1.299
- Author built/tested on 2.1.299; only runs where .claude/principles.md exists.
- One extra model call per code-changing turn.