Help us improve
Share bugs, ideas, or general feedback.
From memory-powermem
Initializes PowerMem for Claude Code after installation. Runs status and init scripts, configures model preferences, handles package source.
npx claudepluginhub oceanbase/powermem --plugin memory-powermemHow this skill is triggered — by the user, by Claude, or both
Slash command
/memory-powermem:initThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Initialize PowerMem for Claude Code.
Explains how claude-mem captures observations, injects context, and where data is stored locally.
Installs, configures, or upgrades ClaudeMemory: checks gem version, runs doctor, inits databases, adds version markers, and reports status.
Integrates mnemonic memory capture and recall into existing Claude Code plugins using sentinel markers for updates, removal, and git rollback. Triggers on integrate/wire/add memory requests.
Share bugs, ideas, or general feedback.
Initialize PowerMem for Claude Code.
Read apps/claude-code-plugin/SETUP.md, section "Installed plugin initialization",
and follow only that section.
Do not run the source/developer setup flow from SETUP.md: do not build hook
binaries, do not stage the plugin, do not run claude plugin marketplace add, do
not run claude plugin install, and do not build the dashboard.
Use the plugin scripts as directed by that section:
scripts/status.shscripts/init.shRemember that scripts/init.sh installs powermem from PyPI by default. If the
user is testing unpublished backend changes, run the script with
POWERMEM_INIT_PACKAGE='powermem @ git+https://github.com/oceanbase/powermem.git@<branch-or-sha>'
instead of using the default PyPI package.
If values are missing, ask only for the missing values and pass them through
POWERMEM_INIT_* environment variables. Never print API keys; mask secrets in
summaries.
If the user is in a network where HuggingFace may be slow or blocked, offer to
run init with POWERMEM_INIT_PRELOAD_MODEL=1 so the script downloads the default
embedding model through ModelScope and bridges it into the HuggingFace cache.