Agent Nurture Framework (OpenClaw Edition)
v1.2.0Agent自我培养框架 — 基于"知识结晶循环"持续成长。 Use when: (1) 需要系统化学习新领域，(2) 从经验中提取可复用技能，(3) 管理技能碎片化，(4) 衡量自身能力成长，(5) 帮助培养新出生的agent。 基于师兄的Agent Nurture Framework (2026-04-04)，...
⭐ 0· 59·0 current·0 all-time
byGuohongbin@guohongbin-git
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
Name/description (Agent Nurture Framework) match the SKILL.md content: guidance for studying, verifying, extracting skills, session reviews, checklists, and file-based memory/skill artifacts. No unrelated credentials, binaries, or installs are requested.
Instruction Scope
SKILL.md instructs the agent to create, update, archive, and delete files (e.g., memory/*.md, session-review-*.md, skill files). That is coherent with a 'nurture' framework, but these file-IO instructions mean the agent will modify workspace state (including deletion/archival rules). Confirm the agent's filesystem sandboxing, backup/restore practices, and limits on where the skill may write or delete to avoid accidental data loss.
Install Mechanism
Instruction-only skill with no install spec and no code files — minimal install risk (nothing downloaded or executed from external sources).
Credentials
No environment variables, credentials, or config paths are requested. The requested actions (file creation/management in memory/ and skill-related files) are proportionate to the stated purpose.
Persistence & Privilege
always:false and default agent invocation settings. The skill does not request permanent presence or modification of other skills' configurations. It does describe writing its own artifacts (memories, session reviews), which is expected for this function.
Assessment
This skill appears coherent and low-risk in that it asks for no credentials or installs. However: (1) it explicitly instructs the agent to write, update, archive, and delete files under memory/ and to create/update skill files — make sure those paths are sandboxed and backed up; (2) confirm operator sign-off workflows are actually enforced (many templates require human confirmation); (3) test in a safe environment to verify the agent's file operations behave as intended and do not touch unrelated data; (4) if you rely on retention of raw session data, ensure archive/delete rules (e.g., '2 weeks unreferenced → archive/delete') align with your retention policy before enabling autonomous execution.Like a lobster shell, security has layers — review code before you run it.
latestvk97923hjsc710kmpgqj1rjpj5s84btpe
License
MIT-0
Free to use, modify, and redistribute. No attribution required.