Evo Memory

This skill appears to do what it claims: it captures signals from your conversations and tool outputs and stores them under ~/.openclaw/workspace/self-evolution/ for later reflection. Before installing: (1) Review the two shell hooks and bootstrap.js to confirm you trust local persistence and that your environment provides the CLAUDE_* variables used to capture inputs/outputs. (2) Be aware the system will persist user utterances and tool error text (potentially sensitive) into local files (principles.md, patterns/*.md, pending.jsonl); decide whether that retention policy and file location are acceptable. (3) Installation requires copying a handler into ~/.openclaw/hooks/ and enabling the hook—this modifies your agent's configuration; only proceed if you trust the skill source. (4) If you proceed, inspect and run the scripts with least privilege (set appropriate file permissions), and optionally run them in a sandboxed account or environment first. (5) If you need higher assurance, request the skill author/source or a signed release; absence of network calls and credentials reduces risk but does not remove privacy implications.