Memory Processor

Enables AI agents to develop independent, evolving personas through organic memory growth, self-reflection, and layered memory management.

This package appears to implement the advertised persona/memory functionality, but there are multiple coherence and security concerns you should address before running it: 1) Do not run as root; change BASE_DIR from /root/.openclaw to a directory you control (or a temporary/isolated path). 2) Review and modify config.py (HOST, PORT, ALLOWED_ORIGINS, API_KEY) so the HTTP API is not publicly open or unauthenticated by default. 3) Inspect requirements.txt for any dependencies that make network calls or download models; decide whether to run in an isolated VM/container. 4) Be aware the service will persist user-provided content to disk (L3 markdown files, SQLite, FAISS index). If that content is sensitive, plan encryption or limited retention. 5) Fix documentation mismatches (example docs show port 9090 but config defaults to 8080). 6) If you need stronger assurance, ask the maintainer for an explanation of the default BASE_DIR choice, a documented authentication model, and full contents of embedding and persona service files to confirm whether they call external APIs. Running this code in a sandboxed environment (container/VM) and performing a manual code review of omitted service files (embedding/persona_service) is recommended before enabling it in any production agent.

Type: OpenClaw Skill Name: memprocessor Version: 0.1.1 The skill bundle is classified as suspicious primarily due to the use of 'pickle.loads' in 'app/core/l1_storage.py', which is a well-known Remote Code Execution (RCE) vulnerability if the cached data is tampered with. Furthermore, 'memory-manager.sh' and 'app/config.py' contain hardcoded paths to the '/root' directory and specific Conda environments, suggesting the application expects to run with elevated privileges and lacks proper environment isolation. While the 'persona' generation and multi-tier memory logic appear consistent with the stated purpose, these high-risk technical choices and broad filesystem assumptions pose a security risk to the host system.