Self-Improving Agent
Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Clau...
v3.0.21
2026/05/01
re-upload
v3.0.16
2026/04/16
暂无更新说明。
v3.0.13
2026/04/03
暂无更新说明。
v3.0.12
2026/04/02
暂无更新说明。
v3.0.8
2026/03/27
暂无更新说明。
v3.0.5
2026/03/17
暂无更新说明。
v3.0.2
2026/03/14
暂无更新说明。
v3.0.1
2026/03/11
暂无更新说明。
v1.0.11
2026/02/22
暂无更新说明。
v1.0.7
2026/02/21
暂无更新说明。
v1.0.5
2026/02/03
暂无更新说明。
v1.0.4
2026/01/31
暂无更新说明。
v1.0.2
2026/01/26
暂无更新说明。
v1.0.1
2026/01/19
暂无更新说明。
v1.0.0
2026/01/05
暂无更新说明。
Security Scan
状态
clean
OpenClaw
gpt-5.5
OpenClaw 分析
This skill persistently records local learning notes and optional reminders, but the behavior is disclosed, purpose-aligned, and not exfiltrating or destructive.
置信度: high
VirusTotal
Type: OpenClaw Skill Name: self-improving-agent Version: 3.0.21 The skill bundle implements a structured 'self-improvement' framework for OpenClaw agents, allowing them to log errors, corrections, and feature requests to local markdown files. It includes utility scripts (e.g., `extract-skill.sh`, `error-detector.sh`) and OpenClaw hooks (`handler.js`) designed to automate the capture of learnings and provide contextual reminders to the agent. While the skill facilitates the modification of agent instruction files like `SOUL.md` and `AGENTS.md`, it includes explicit safety guidelines advising the agent not to log secrets, tokens, or sensitive environment variables. The scripts utilize standard shell commands and include basic security checks, such as path traversal prevention, aligning entirely with the stated purpose of continuous workflow optimization.