Review-gate OpenClaw memory hygiene with openclaw-mem

Pack trusted context and review memory writes before long OpenClaw sessions drift or accumulate low-quality memory.

Prerequisites

OpenClaw, SQLite, local filesystem access

Installation

Use the upstream install or setup path that matches your environment:

• pip install openclaw-context-pack
• git clone https://github.com/phenomenoner/openclaw-mem.git
• uv sync --locked
• uv run --python 3.13 --frozen -- \

Requirements and caveats from upstream:

• python -m venv .venv
• python benchmarks/trust_policy_synthetic_proof.py --json

Basic usage or getting-started notes:

• Run the synthetic proof: Trust-policy synthetic proof

• Pack: run pack to get a bounded bundle_text and context_pack (schema: openclaw-mem.context-pack.v1), with citations, trust policy, and trace receipts.

• openclaw-mem self-curator verify --receipt .state/self-curator/apply-runs/<run>/apply-receipt.json --json

• Source: https://github.com/phenomenoner/openclaw-mem

• Extracted from upstream docs: https://raw.githubusercontent.com/phenomenoner/openclaw-mem/HEAD/README.md

Documentation

• https://phenomenoner.github.io/openclaw-mem/

Source

• Agent Skill Exchange