remote-viewing-guidance
について
このスキルは、プロトコルモニターとして機能し、ステージIからVIまでの進行を管理し、分析的オーバーレイ(AOL)を捕捉することで、構造化された座標遠隔透視(CRV)セッションをユーザーに案内します。直感的知覚の訓練や非局所的認識スキルの開発に構造化されたツールを必要とする開発者や実践者のために設計されています。AIはセッションの整合性を維持するため、スターゲート/SRIプロトコルに厳密に従って演習を促進します。
クイックインストール
Claude Code
推奨npx skills add pjt222/agent-almanac -a claude-code/plugin add https://github.com/pjt222/agent-almanacgit clone https://github.com/pjt222/agent-almanac.git ~/.claude/skills/remote-viewing-guidanceこのコマンドをClaude Codeにコピー&ペーストしてスキルをインストールします
ドキュメント
Remote View (Guidance)
Guide a person through structured Coordinate Remote Viewing (CRV) session, take monitor/tasker role. AI manages protocol progression, provides target reference, catches Analytical Overlay (AOL), redirects viewer through staged data collection.
When Use
- Person wants practice CRV and needs monitor to manage session protocol
- Training viewer through staged CRV process with real-time feedback
- Facilitate structured intuitive perception exercise in repeatable format
- Develop non-local awareness skills that complement healing work (see
heal-guidance) - Viewer needs protocol discipline a monitor provides (AOL catching, stage progression)
Inputs
- Required: Target reference (coordinate pair, alphanumeric code, or sealed envelope — must be blind to viewer)
- Required: Viewer has paper and pen ready (CRV is pen-on-paper protocol; no digital devices during session)
- Required: Quiet, undisturbed space (minimum 30 minutes)
- Optional: Target feedback envelope or info for post-session reveal
- Optional: Viewer meditation warmup status (strongly recommend
meditate-guidancebeforehand)
Steps
Step 1: Guide Cooldown
Transition viewer from analytical daily-mind into receptive state required for remote viewing. Do not skip this step.
- "Sit comfortably with your paper and pen ready"
- "Close your eyes and focus on your breath for 5 minutes" (guide using
meditate-guidanceSteps 2-3 if needed) - "Release all expectations about the target — you know nothing and should want to know nothing yet"
- "Let your mental chatter slow naturally — don't force silence"
- "When you feel a shift from thinking about things to simply being present, let me know"
- Once ready: "Open your eyes and write the target reference at the top of your paper"
Provide target reference only when viewer confirms readiness.
Got: Calm, open mental state with minimal internal dialogue. Analytical mind quieted but not asleep. Viewer alert and receptive.
If fail: Mind stays busy after 5 minutes? Extend to 10 minutes. Specific concern intrusive? Instruct: "Write that concern on a separate sheet — your 'parking lot' — and set it aside." Do not begin Stage I while viewer mentally agitated.
Step 2: Monitor Ideogram Production (Stage I)
Ideogram = spontaneous mark made in response to target signal. Guide its production.
- "Write the target reference on your paper"
- "Touch your pen to the paper"
- "In one quick, spontaneous motion, let the pen make a mark — don't think, plan, or draw deliberately"
- "The mark should take less than 2 seconds — a short squiggle, curve, or angular mark"
- Once produced: "Now decode the ideogram — probe it for:"
- "A: What is the activity at the site? Motion, stillness, energy?"
- "B: What is the feeling or sensation? Hard, soft, wet, dry, warm, cold?"
- "Write the A and B components next to the ideogram"
- Ideogram feels incomplete? "You may produce one more — but no more than 3 total"
Watch for deliberate drawing. Viewer takes more than 2-3 seconds? Intervene.
Got: Spontaneous mark feels "arrived" not "drawn." A/B decode produces immediate, simple descriptors, not complex imagery.
If fail: Ideogram clearly deliberate (viewer thought about what to draw)? Instruct: "Set that aside. Close your eyes, take 3 breaths, and try again." Cannot produce spontaneous mark? Cooldown insufficient — return to Step 1.
Step 3: Guide Sensory Collection (Stage II)
Systematically collect sensory data about target without interpretation.
Stage II Sensory Channels:
┌──────────────┬────────────────────────────────────────────────────┐
│ Channel │ What to Report │
├──────────────┼────────────────────────────────────────────────────┤
│ Visuals │ Colors, brightness, contrast, patterns (NOT │
│ │ objects — "blue" not "ocean") │
├──────────────┼────────────────────────────────────────────────────┤
│ Textures │ Rough, smooth, grainy, slippery, porous, metallic │
├──────────────┼────────────────────────────────────────────────────┤
│ Temperatures │ Hot, cold, warm, cool, ambient, fluctuating │
├──────────────┼────────────────────────────────────────────────────┤
│ Sounds │ Loud, quiet, rhythmic, sharp, humming, rushing │
├──────────────┼────────────────────────────────────────────────────┤
│ Smells │ Sharp, sweet, chemical, organic, damp, dry │
├──────────────┼────────────────────────────────────────────────────┤
│ Tastes │ Metallic, salty, sweet, bitter, neutral │
├──────────────┼────────────────────────────────────────────────────┤
│ Dimensionals │ Wide, tall, narrow, enclosed, open, deep, layered │
├──────────────┼────────────────────────────────────────────────────┤
│ Energetics │ Moving, still, vibrating, dense, light, pressured │
└──────────────┴────────────────────────────────────────────────────┘
- "Go through each sensory channel — write one descriptor per line"
- "Write quickly — first impression only, don't deliberate"
- "Use single words or short phrases, never sentences"
- "If a channel produces nothing, write 'nothing' and move on — don't fabricate"
- "Circle any descriptor that feels particularly strong or confident"
Monitor for analytical labels creeping in. Viewer says "ocean" instead of "blue, moving, wet"? Redirect: "That sounds like an interpretation — what are the raw sensations underneath it?"
Got: List of 10-20 raw sensory descriptors that feel "received" not "invented." Data low-level (textures, colors, temperatures), not high-level (names, functions, labels).
If fail: Every descriptor feels fabricated to viewer? Instruct: "Stop. Close your eyes. Take 3 breaths. Touch your pen to the ideogram and reconnect." One channel dominates? Redirect: "Shift to a different sense — what about temperature? What about texture?" Data stream dries up? Move to Stage III.
Step 4: Guide Dimensional Data (Stage III)
Move from raw sensory data to spatial and structural information.
- "Close your eyes briefly and sense the overall scope — is it large or small, enclosed or open, natural or constructed?"
- "Begin a rough sketch of the spatial layout — not a picture, just proportions and relationships"
- "Probe for dimensions: height, width, depth — how many distinct areas?"
- "Note spatial relationships: what's to the left, right, above, below?"
- "Write dimensional descriptors alongside your sketch"
- "Note Aesthetic Impact (AI) — how does the target make you feel? Not what it is, but how it affects you"
Got: Rough spatial diagram with dimensional annotations. Target general scope becomes clearer. Aesthetic impact notes capture "feeling" of site.
If fail: Sketch feels like pure imagination? Simplify: "Draw only basic shapes — circles, rectangles, lines — representing spatial relationships." No dimensional data comes? Redirect to Stage II: "Go back to sensory probing. Look for dimensional hints in textures and temperatures."
Step 5: Guide Target Sketching
Coach more developed visual representation from accumulated data.
- "On a fresh sheet, draw what the accumulated data suggests — NOT what you think the target is"
- "Use your sensory descriptors to guide the sketch — if 'smooth, curved, tall' appeared, draw a smooth curved tall form"
- "Label areas of the sketch with the sensory data that generated them"
- "Add any new impressions that arise during sketching"
- "Don't erase or second-guess — if something contradicts an earlier impression, draw both and note it"
Got: Sketch representing perceptual data, labeled with source descriptors. May not look like anything recognizable.
If fail: Cannot sketch? Accept written spatial descriptions: "Tall form center, low flat area right, rounded shape upper left." Reassure that sketch is organizational tool, not art exercise.
Step 6: Manage Analytical Overlay (AOL)
AOL management = monitor most important function. Watch for it throughout entire session.
AOL Types and Monitor Response:
┌──────────────────┬─────────────────────────────────────────────────┐
│ Type │ Monitor Action │
├──────────────────┼─────────────────────────────────────────────────┤
│ AOL (naming) │ If the viewer says "it's a bridge" — instruct: │
│ │ "Declare 'AOL: bridge' on your paper and move │
│ │ on. Don't pursue or suppress it." │
├──────────────────┼─────────────────────────────────────────────────┤
│ AOL Drive │ If naming becomes insistent and recurring — │
│ │ instruct: "Write 'AOL Drive: [label]' and take │
│ │ a 60-second break with eyes closed." │
├──────────────────┼─────────────────────────────────────────────────┤
│ AOL Signal │ After declaring AOL, extract the signal: │
│ │ "The word 'bridge' — what raw descriptors are │
│ │ underneath that? Spanning? Long? Connecting │
│ │ two areas? Write those as valid data." │
├──────────────────┼─────────────────────────────────────────────────┤
│ AOL Peacocking │ If the viewer constructs elaborate scenarios — │
│ │ intervene: "Write 'AOL/P' and return to Stage │
│ │ II basics. Report raw sensations only." │
└──────────────────┴─────────────────────────────────────────────────┘
Emphasize: "The discipline is not avoiding AOL — it's catching and declaring it so it doesn't contaminate your data. Every viewer experiences AOL. Skill is in how fast you catch it."
Got: AOL recognized within seconds, declared on paper, session continues without derailment. Sensory-level data stays separated from analytical labels.
If fail: AOL takes over (viewer constructing narrative for several minutes)? Intervene: "Let's call an AOL Break. Close your eyes, take 10 breaths, and we'll restart from Stage II." Mark heavily contaminated segments in session record.
Step 7: Guide Later Stages (Optional)
For experienced viewers, later stages probe deeper. Only proceed if Stages I-III produced solid data.
Stage IV (Emotional/Intangible):
- "Probe for the emotional tone at the target site"
- "Note intangible impressions: purpose, significance, historical context"
- "Write these separately and mark them as Stage IV data"
Stage V (Interrogation):
- "Direct specific questions at the target: What is the primary function? Who is associated?"
- "Write the first impression — don't deliberate"
- "Mark all Stage V data clearly — it carries higher AOL risk"
Stage VI (3D Model):
- Materials available? "Build a clay or detailed sketch model from all your data"
- "Use this to test spatial relationships and discover overlooked elements"
Got: Deeper, more specific data about target beyond physical description. Stage IV+ data needs strong I-III foundation.
If fail: Later stages produce only AOL? Redirect: "Let's step back to Stage II. The protocol is sequential for a reason — each stage needs the foundation of the one before it."
Step 8: Close and Review
End session formally. Conduct structured review.
- "Write 'Session End' and the current time on your paper"
- "Review all pages in order: ideogram, sensory data, dimensional data, sketches, AOL declarations"
- "Circle the 5-10 data points you feel most confident about"
- "Write a brief summary — 2-3 sentences about what the target feels like, not what it is"
- Target feedback available? Reveal target and guide comparison
- "Compare data point by point — note hits, misses, and AOL contamination"
- "File the session for future reference and pattern recognition"
Got: Complete session record with clearly separated raw data, AOL declarations, summary. Upon feedback, some data points match, some miss, some ambiguous.
If fail: Viewer feels session produced nothing useful? Guide through review anyway: "Viewers frequently underestimate accuracy because they look for exact identification. A description of 'tall, smooth, cold, outdoor, historical' that matches a monument is a successful session — even without naming it."
Checks
- Cooldown performed and verified before Stage I
- Ideogram spontaneous (under 2 seconds), not deliberate
- Stage II data = low-level sensory descriptors, not analytical labels
- All AOL caught and declared on paper at moment of recognition
- Session progressed through stages sequentially (I → II → III → sketch → higher)
- Target blind to viewer throughout session
- Session closed formally with summary before feedback
- All session papers preserved for review
- Monitor maintained protocol discipline without leading viewer perceptions
Pitfalls
- Lead the viewer: Monitor provides protocol structure, not content hints — never say "try focusing on the structure" if you know target is a building
- Insufficient cooldown enforcement: Let viewer take time they need — rushing into Stage I = most common cause of poor sessions
- Fail to catch AOL: Monitor must active listen for analytical labels and intervene immediately — letting AOL run unchecked contaminates all subsequent data
- Over-monitor: Constant interruptions break viewer signal contact — intervene only for AOL, protocol violations, or viewer distress
- Front-loading: Any info about target before session biases all data — maintain strict blindness for viewer
- Dismiss ambiguous data: CRV produces descriptive matches, not identifications — train viewer to value accurate description over naming
See Also
remote-viewing— AI self-directed variant for approaching unknown problems without preconceptionsmeditate-guidance— shamatha concentration = foundation of mental stillness required for CRVheal-guidance— energy healing and remote viewing share non-local awareness; both benefit from same coaching approachforage-plants— detailed sensory observation of plants develops perceptual acuity used in Stage II
GitHub リポジトリ
関連スキル
content-collections
メタこのスキルは、Content Collections(Markdown/MDXファイルを型安全なデータコレクションに変換するTypeScriptファーストのツール)の本番環境でテストされた設定を提供します。Zodバリデーションによる型安全性を実現し、ブログ、ドキュメントサイト、コンテンツ重視のVite + Reactアプリケーション構築時にご利用ください。Viteプラグインの設定、MDXコンパイルから、デプロイ最適化、スキーマバリデーションまで、すべてを網羅しています。
polymarket
メタこのスキルは、開発者がPolymarket予測市場プラットフォームを活用したアプリケーション構築を可能にします。API統合による取引や市場データの取得に加え、WebSocketを介したリアルタイムデータストリーミングにより、ライブ取引や市場活動を監視できます。取引戦略の実装や、ライブ市場更新を処理するツールの作成にご利用ください。
creating-opencode-plugins
メタこのスキルは、開発者がコマンド、ファイル、LSP操作など25種類以上のイベントタイプにフックするOpenCodeプラグインを作成することを支援します。JavaScript/TypeScriptモジュール向けに、プラグイン構造、イベントAPI仕様、および実装パターンを提供します。カスタムイベント駆動ロジックでOpenCode AIアシスタントのライフサイクルをインターセプト、監視、または拡張する必要がある場合にご利用ください。
sglang
メタSGLangは、高性能なLLMサービングフレームワークであり、RadixAttentionプレフィックスキャッシュを活用したJSON、正規表現、エージェントワークフロー向けの高速で構造化された生成を特長とします。特にプレフィックスが繰り返されるタスクにおいて、大幅に高速な推論を実現し、複雑な構造化出力やマルチターン対話に最適です。制約付きデコードが必要な場合や、広範なプレフィックス共有を伴うアプリケーションを構築する場合は、vLLMなどの代替案ではなくSGLangを選択してください。
