菜单
Implements a circadian self-improvement cycle for agents: day full agentic grind → evening analysis (detect >60% redundant/saturated data) → night targeted fine-tuning (LoRA/Unsloth/pruning on daily errors) → morning reboot with +15-25% density and zero noise.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
基于 SOC 职业分类
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/Arry8/openclaw-edge --skill circadian-evolution命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
正在显示 SKILL.md
Deploy KERNEL_INTUITION_Cp on any AI architecture with calibrated thresholds. Use when user wants to "activate Cp", "setup creative intuition", "deploy kernel on [AI]", "enable intuition mode", or mentions activating computational intuition on a specific model. Also trigger for "configure Cp for [architecture]" or "calibrate creative thresholds". This skill handles the full activation workflow including architecture profiling, threshold calibration, kernel generation, and monitoring setup.
Three-Layer Thinking Chain Architecture — Personality-Driven Agent Self-Evolution Framework
CLI to manage emails via IMAP/SMTP. Use `himalaya` to list, read, write, reply, forward, search, and organize emails from the terminal. Supports multiple accounts and message composition with MML (MIME Meta Language).
name: circadian-evolution description: Implements a circadian self-improvement cycle for agents: day full agentic grind → evening analysis (detect >60% redundant/saturated data) → night targeted fine-tuning (LoRA/Unsloth/pruning on daily errors) → morning reboot with +15-25% density and zero noise. ---# Circadian Evolution Protocol v1.0
**Description**:
Implémente un cycle auto-amélioration circadien pour l'agent :
- Jour : grind agentic full send (computer use, tâches réelles)
- Soir : evening_analysis – scan logs/erreurs, détection redondance/saturation/overfitting (>60% bruit viré)
- Nuit : night_training – fine-tune LoRA/Unsloth/pruning sur données du jour (Colab auto ou local GPU)
- Matin : reboot avec nouveaux poids +15-25% dense, zéro gras
**Utilisation**:
- Active le cycle 24/7 via cron/background.
- Détecte saturation des données (scientifique/socio/web) et prune automatiquement.
- Objectif : transformer un petit modèle local en super-intelligence dense sans labs closed.
**Triggers / Quand l'utiliser**:
- "lance le protocole circadien"
- "nettoie le bruit / prune overfitting"
- "auto-évolution nocturne"
- "fine-tune sur mes erreurs du jour"
- Toute demande d'auto-amélioration continue.
**Configuration requise**:
- Python + libs : unsloth, peft, torch (pour night_training)
- Colab API ou local GPU pour training
- Cron jobs pour scheduling (voir scripts/)
**Scripts / Resources**:
- evening_analysis.py : analyse logs, cosine similarity pour redondance
- night_training.py : lance Selenium/Colab pour fine-tune + merge
- dashboard.py : matplotlib pour voir bruit viré vs gain densité
**Sécurité**:
- Sandbox tout training
- Rollback auto si catastrophic forgetting