MemoryLane

Setup

Run bash install.sh from the repo root to initialize .memorylane/ and make the CLI executable.

Ensure .claude/settings.json points at .claude/hooks/*.py so automatic context injection and learning run.

Verify the install with:

python3 src/cli.py status

Core commands

Check status and savings:

python3 src/cli.py status

Recall memories:

python3 src/cli.py recall "<query>"

View insights and costs:

python3 src/cli.py insights
python3 src/cli.py costs

Learn from history:

python3 src/learner.py initial

Memory management

List and curate memories:

python3 src/cli.py curate --list
python3 src/cli.py curate --apply '<JSON>'

Manage an individual memory:

python3 src/cli.py memory get <id>
python3 src/cli.py memory update <id> --content "New content"
python3 src/cli.py memory delete <id>

Server

Start or stop the sidecar:

python3 src/server.py start
python3 src/server.py status
python3 src/server.py stop

Notes

Do not edit .memorylane/memories.json directly; use the CLI.

Context rot guard: injected context is capped to a safe fraction of the model window via context_rot.* config keys.