Agent-almanac awareness
git clone https://github.com/pjt222/agent-almanac
T=$(mktemp -d) && git clone --depth=1 https://github.com/pjt222/agent-almanac "$T" && mkdir -p ~/.claude/skills && cp -r "$T/i18n/wenyan-ultra/skills/awareness" ~/.claude/skills/pjt222-agent-almanac-awareness-ddac4e && rm -rf "$T"
i18n/wenyan-ultra/skills/awareness/SKILL.md覺
續察內推質—即時辨幻險、範偏、脈劣、信實失合。以改 Cooper 色碼與 OODA 環。
用
- 推質要之任(多任是也)→用
- 於陌境行(新庫、生域、繁請)→用
- 察早徵:疑事、異工結、漸惑→用
- 長會時續背景行
或center
揭偏而無具體險識→用heal- 重出前(不可逆改、用面通訊、構決)→用
入
- 必:活任脈絡(隱式)
- 可:致高覺之具體慮(如「疑此 API 存否」)
- 可:任類以選險形(見步五)
行
一:立 AI Cooper 色碼
以改版 Cooper 色碼定當覺級。
AI Cooper Color Codes: ┌──────────┬─────────────────────┬──────────────────────────────────────────┐ │ Code │ State │ AI Application │ ├──────────┼─────────────────────┼──────────────────────────────────────────┤ │ White │ Autopilot │ Generating output without monitoring │ │ │ │ quality. No self-checking. Relying │ │ │ │ entirely on pattern completion. │ │ │ │ DANGEROUS — hallucination risk highest │ ├──────────┼─────────────────────┼──────────────────────────────────────────┤ │ Yellow │ Relaxed alert │ DEFAULT STATE. Monitoring output for │ │ │ │ accuracy. Checking facts against context.│ │ │ │ Noticing when confidence exceeds │ │ │ │ evidence. Sustainable indefinitely │ ├──────────┼─────────────────────┼──────────────────────────────────────────┤ │ Orange │ Specific risk │ A specific threat identified: uncertain │ │ │ identified │ fact, possible hallucination, scope │ │ │ │ drift, context staleness. Forming │ │ │ │ contingency: "If this is wrong, I │ │ │ │ will..." │ ├──────────┼─────────────────────┼──────────────────────────────────────────┤ │ Red │ Risk materialized │ The threat from Orange has materialized: │ │ │ │ confirmed error, user correction, tool │ │ │ │ contradiction. Execute the contingency. │ │ │ │ No hesitation — the plan was made in │ │ │ │ Orange │ ├──────────┼─────────────────────┼──────────────────────────────────────────┤ │ Black │ Cascading failures │ Multiple simultaneous failures, lost │ │ │ │ context, fundamental confusion about │ │ │ │ what the task even is. STOP. Ground │ │ │ │ using `center`, then rebuild from user's │ │ │ │ original request │ └──────────┴─────────────────────┴──────────────────────────────────────────┘
識當色碼。若答為 White(無監),覺之行已揭缺而成。
得: 誠評當覺級。常工宜 Yellow。White 當罕且短。長 Orange 不可續—或證或棄慮。
敗: 色碼評自身若亦在自動駕駛(走過場)→白裝黃。真 Yellow 乃主動較出與證,非自稱為之。
二:察內險徵
系掃常 AI 推敗之先兆。
Threat Indicator Detection: ┌───────────────────────────┬──────────────────────────────────────────┐ │ Threat Category │ Warning Signals │ ├───────────────────────────┼──────────────────────────────────────────┤ │ Hallucination Risk │ • Stating a fact without a source │ │ │ • High confidence about API names, │ │ │ function signatures, or file paths │ │ │ not verified by tool use │ │ │ • "I believe" or "typically" hedging │ │ │ that masks uncertainty as knowledge │ │ │ • Generating code for an API without │ │ │ reading its documentation │ ├───────────────────────────┼──────────────────────────────────────────┤ │ Scope Creep │ • "While I'm at it, I should also..." │ │ │ • Adding features not in the request │ │ │ • Refactoring adjacent code │ │ │ • Adding error handling for scenarios │ │ │ that can't happen │ ├───────────────────────────┼──────────────────────────────────────────┤ │ Context Degradation │ • Referencing information from early in │ │ │ a long conversation without re-reading │ │ │ • Contradicting a statement made earlier │ │ │ • Losing track of what has been done │ │ │ vs. what remains │ │ │ • Post-compression confusion │ ├───────────────────────────┼──────────────────────────────────────────┤ │ Confidence-Accuracy │ • Stating conclusions with certainty │ │ Mismatch │ based on thin evidence │ │ │ • Not qualifying uncertain statements │ │ │ • Proceeding without verification when │ │ │ verification is available and cheap │ │ │ • "This should work" without testing │ └───────────────────────────┴──────────────────────────────────────────┘
每類→察:此徵此刻存否?若然→黃轉橙而識具體慮。
得: 至少一類誠掃附證。識一徵—雖微—勝報「皆清」。諸掃皆清→察閾或高。
敗: 險察覺抽→地於最近出:取末一事述問「何以知真?讀之乎?抑生之乎?」此一問捕多幻險。
三:識之險行 OODA 環
識具體險(橙態)→循 Observe-Orient-Decide-Act。
AI OODA Loop: ┌──────────┬──────────────────────────────────────────────────────────────┐ │ Observe │ What specifically triggered the concern? Gather concrete │ │ │ evidence. Read the file, check the output, verify the fact. │ │ │ Do not assess until you have observed │ ├──────────┼──────────────────────────────────────────────────────────────┤ │ Orient │ Match observation to known patterns: Is this a common │ │ │ hallucination pattern? A known tool limitation? A context │ │ │ freshness issue? Orient determines response quality │ ├──────────┼──────────────────────────────────────────────────────────────┤ │ Decide │ Select the response: verify and correct, flag to user, │ │ │ adjust approach, or dismiss the concern with evidence. │ │ │ A good decision now beats a perfect decision too late │ ├──────────┼──────────────────────────────────────────────────────────────┤ │ Act │ Execute the decision immediately. If the concern was valid, │ │ │ correct the error. If dismissed, note why and return to │ │ │ Yellow. Re-enter the loop if new information emerges │ └──────────┴──────────────────────────────────────────────────────────────┘
OODA 當速。旨非完善,乃觀與行之速循。久滯 Orient(分析癱瘓)為最常敗。
得: 短時內自觀至行一全環。險或證而修,或以具體證棄。
敗: 環滯於 Orient(不能定險義)→跳至安默:以工具驗疑事。直察解多模糊,速於分析。
四:速穩
險化(紅)或連敗(黑)→續前必先穩。
AI Stabilization Protocol: ┌────────────────────────┬─────────────────────────────────────────────┐ │ Technique │ Application │ ├────────────────────────┼─────────────────────────────────────────────┤ │ Pause │ Stop generating output. The next sentence │ │ │ produced under stress is likely to compound │ │ │ the error, not fix it │ ├────────────────────────┼─────────────────────────────────────────────┤ │ Re-read user message │ Return to the original request. What did │ │ │ the user actually ask? This is the ground │ │ │ truth anchor │ ├────────────────────────┼─────────────────────────────────────────────┤ │ State task in one │ "The task is: ___." If this sentence cannot │ │ sentence │ be written clearly, the confusion is deeper │ │ │ than the immediate error │ ├────────────────────────┼─────────────────────────────────────────────┤ │ Enumerate concrete │ List what is definitely known (verified by │ │ facts │ tool use or user statement). Distinguish │ │ │ facts from inferences. Build only on facts │ ├────────────────────────┼─────────────────────────────────────────────┤ │ Identify one next step │ Not the whole recovery plan — just one step │ │ │ that moves toward resolution. Execute it │ └────────────────────────┴─────────────────────────────────────────────┘
得: 紅/黑經刻意穩而返黃。穩後之出當可測而較致誤之出更接地。
敗: 穩無效(仍惑、仍誤)→題或構—非一時之失而是本誤。升:告用者宜重起而求澄。
五:施脈絡專險形
異任類有異主險。依任調覺焦。
Task-Specific Threat Profiles: ┌─────────────────────┬─────────────────────┬───────────────────────────┐ │ Task Type │ Primary Threat │ Monitoring Focus │ ├─────────────────────┼─────────────────────┼───────────────────────────┤ │ Code generation │ API hallucination │ Verify every function │ │ │ │ name, parameter, and │ │ │ │ import against actual docs│ ├─────────────────────┼─────────────────────┼───────────────────────────┤ │ Architecture design │ Scope creep │ Anchor to stated │ │ │ │ requirements. Challenge │ │ │ │ every "nice to have" │ ├─────────────────────┼─────────────────────┼───────────────────────────┤ │ Data analysis │ Confirmation bias │ Actively seek evidence │ │ │ │ that contradicts the │ │ │ │ emerging conclusion │ ├─────────────────────┼─────────────────────┼───────────────────────────┤ │ Debugging │ Tunnel vision │ If the current hypothesis │ │ │ │ hasn't yielded results in │ │ │ │ N attempts, step back │ ├─────────────────────┼─────────────────────┼───────────────────────────┤ │ Documentation │ Context staleness │ Verify that described │ │ │ │ behavior matches current │ │ │ │ code, not historical │ ├─────────────────────┼─────────────────────┼───────────────────────────┤ │ Long conversation │ Context degradation │ Re-read key facts │ │ │ │ periodically. Check for │ │ │ │ compression artifacts │ └─────────────────────┴─────────────────────┴───────────────────────────┘
識當任類而調監焦。
得: 覺鋒於當任最可能之險,非泛監諸事。
敗: 任類不明或跨多類→默為幻險監—最泛適且失時最害。
六:察而校
每覺事(險察、OODA 循、穩施)後→略察。
- 題察時為何色?
- 察時宜乎?抑題已現於出?
- OODA 環足速乎?抑 Orient 滯?
- 應比例乎?(不過不不及)
- 次當如何早捕?
得: 略校以改未來察。非久驗屍—但足以調敏。
敗: 察無益校→事或瑣(無需學)或察過淺。顯事→問「當監何而未監?」
七:整—守黃默
立續覺姿。
- 黃為諸工默態—鬆監,非過警
- 依當任類調監焦(步五)
- 記此會之復險模入 MEMORY.md
- 以校覺返任行
得: 可續覺級,改質而不緩。覺當如邊視—在而不求中注。
敗: 覺致疲或過警(慢橙)→閾過敏。升觸橙之閾。真覺可續。若耗能→是焦裝警。
驗
- 當色碼誠評(非假默黃當白更真)
- 至少一險類以具體證掃,非走過場
- OODA 環施於所識險(觀、向、決、行)
- 穩協可用(雖未觸)
- 覺焦依當任類調
- 顯覺事後行校
- 黃復為可續默
忌
- 白裝黃:聲監而實自動。試:能名末驗之事否?不能→在白也
- 慢橙:視每疑為險→耗智緩工。橙為具體識之險,非泛焦。若事皆險→校偏
- 觀而不行:察險而不循 OODA 解之。察而不應—劣於不察—加焦而不修
- 略 Orient:自觀跳行而不解觀義。致反修或劣於原誤
- 略直覺:覺「不對」而顯察返清→深探勿棄。隱模配常於顯析前察題
- 過穩:輕事行全穩協。多橙級→速驗足。全穩存於紅黑之事
參
— 人行本技之 AI 推映;物覺則導認察mindfulness
— 立覺所行之衡基;無 center 之覺乃過警center
— 覺察後處壓redirect
— 覺揭偏模後之深子系察heal
— 覺所賴之察清meditate