OpenClaw Memory Fusion

Build and operate a unified OpenClaw memory system that combines native Markdown memory, project memory, structured event memory, migration from older PROJEC...

This skill is largely coherent with its stated purpose, but review the full Python script before running it on your primary workspace because the provided script was truncated and I couldn't fully audit it. Specific actions you should consider: (1) run the commands in a copy of your workspace or a test environment first (use the default dry-run behavior); (2) inspect the complete scripts/openclaw_memory_fusion.py for any network calls, external-provider code paths, or credential use (especially sections that reference models/providers); (3) confirm you are comfortable with the tool reading and writing files under ~/.openclaw/workspace and creating backups under memory_fusion/checkpoints; (4) avoid supplying any unrelated credentials to the script and check whether it prompts for or expects cloud/model keys before enabling non-local providers. If you want, provide the remainder of the script (truncated portion) and I can re-evaluate with higher confidence.

Type: OpenClaw Skill Name: openclaw-memory-fusion-core Version: 1.0.1 The skill is a comprehensive utility designed to unify and migrate OpenClaw memory systems into a structured layout. It includes features for auto-capturing events from daily logs, detecting 'drift' in project status, and configuring the OpenClaw memorySearch system. While the script (openclaw_memory_fusion.py) performs high-privilege actions such as modifying the openclaw.json configuration and performing extensive file operations across the workspace, these behaviors are explicitly documented, aligned with the skill's purpose, and protected by safety mechanisms like dry-run defaults and a manifest-based rollback system. No indicators of data exfiltration, unauthorized execution, or malicious prompt injection were found.