← 返回
未分类 中文

Hui-Yi

Trigger for cold-memory recall and archive work under memory/cold/. Use for older low-frequency context, historical continuity, resurfacing, cooling, rebuild...
触发冷记忆召回与归档(memory/cold/),仅用于明确请求旧低频上下文、历史连续性等。
fuetsui fuetsui 来源
未分类 clawhub v1.2.11 4 版本 99872.6 Key: 无需
★ 1
Stars
📥 764
下载
💾 1
安装
4
版本
#latest

概述

Hui Yi

Hui Yi manages the cold reinforcement layer under memory/cold/.

Core rule:

Repeatedly reactivated information deserves reinforcement first. Ebbinghaus sets the pace, not the sole trigger.

Use Hui Yi when

  • older low-frequency context would materially improve the current answer
  • the user asks what was done before, asks to recall/archive something, or wants historical continuity
  • a reusable lesson, decision, troubleshooting result, or stable background note should be preserved in memory/cold/
  • durable content from daily notes should be cooled into cold memory
  • cold-memory notes, metadata, or retrieval quality need maintenance

Do not use Hui Yi when

  • the content is today's transient note → memory/YYYY-MM-DD.md
  • the content is a stable high-frequency fact → MEMORY.md
  • the content is tooling, machine path, or environment setup → TOOLS.md
  • the content is a fresh mistake or still-unvalidated lesson → .learnings/
  • the content contains secrets, tokens, or passwords

Boundary

OpenClaw primary memory handles:

  • current chat continuity
  • recent daily notes
  • stable high-frequency facts
  • tooling and environment notes
  • fresh learnings

Hui Yi handles:

  • low-frequency, high-value knowledge under memory/cold/
  • historical context that keeps resurfacing across real conversations
  • durable experience, decisions, and troubleshooting notes that should not pollute primary memory

Files and scripts

Cold-memory area:

  • memory/cold/index.md
  • memory/cold/tags.json
  • memory/cold/retrieval-log.md
  • memory/cold/_template.md
  • memory/cold/schedule.json
  • memory/heartbeat-state.json

CLI entry scripts:

  • scripts/create.py
  • scripts/validate.py
  • scripts/search.py
  • scripts/rebuild.py
  • scripts/decay.py
  • scripts/cool.py
  • scripts/review.py
  • scripts/scheduler.py
  • scripts/install_hook.py
  • scripts/smoke_test.py
  • scripts/scrub_metadata.py

Core modules:

  • core/signal_detect.py
  • core/signal_pipeline.py
  • core/openclaw_signal_hook.py
  • core/openclaw_runtime_probe.py

Disclosure

  • Scripts create and modify persistent local files, but only under memory/cold/

and memory/heartbeat-state.json. No network access, no credential access.

  • The hook is optional and twice opt-in: scripts/install_hook.py copies files

into hooks/hui-yi-signal-hook/, and only --enable flips the two

hooks.internal.* keys in openclaw.json (after printing a disclosure; it

refuses to create a missing config file).

  • The enabled hook persists session signals as activation counters plus truncated

sha256 session fingerprints — never raw session keys. Its hook.log by default

contains no message bodies and no raw user/chat ids (set HUI_YI_HOOK_DEBUG=1

for verbose diagnostics) and rotates at 256 KB.

  • The cooling playbook's routing table (MEMORY.md, TOOLS.md, .learnings/,

AGENTS.md, SOUL.md) is agent guidance; Hui-Yi scripts never write those files.

Sanity check

python3 scripts/smoke_test.py

版本历史

共 4 个版本

  • v1.2.11 当前
    2026-06-11 23:29
  • v1.2.10
    2026-06-07 12:19
  • v1.2.9
    2026-05-08 12:31 安全 安全
  • v1.2.5
    2026-05-03 04:20 安全 安全

安全检测

腾讯云安全 (Keen)

队列中

腾讯云安全 (Sanbu)

队列中

🔗 相关推荐

ai-agent

Find Skills

guipi888
场景驱动+关键词双模式技能发现工具。当用户用自然语言描述场景/需求(如"我想做一个海报""帮我分析股票"),或明确说"安装技能/find skills/找个skill"时,自动从官方内置、本地已安装、SkillHub、虾评、GitHub、C
★ 1,479 📥 542,473
ai-agent

Self-Improving + Proactive Agent

ivangdavila
自我反思+自我批评+自我学习+自组织记忆。智能体评估自身工作、发现错误并持续改进。
★ 1,401 📥 323,436
ai-agent

self-improving agent

pskoett
捕获经验教训、错误及修正内容,以实现持续改进。适用于以下场景:(1)命令或操作意外失败;(2)用户纠正Claude(如“不,那不对……”“实际上……”);(3)用户请求的功能不存在;(4)外部API或工具出现故障;(5)Claude发现自身
★ 4,113 📥 835,182