CRV Program

Add new target

This is a tool for practicing Controlled Remote Viewing (CRV) — a structured, blind session methodology. You're given (or generate) a blind target code with no other information, then work through a fixed sequence of stages — ideogram, sensory impressions, structured signals, sketching, and a summary — before revealing what the target actually was and rating how close you got. Every session is saved to your Log so you can track hit rate over time.

On the 👁 Session tab you have two options: click 🔀 New random target to pull a blind code from your own 🎯 Targets pool, or if someone else is running the session for you, have them hand you just a target number under "Someone else gives you target number." Every other field there (label, URL, type, coordinates) is optional and stays hidden until reveal — the blind code is the only thing you actually need to start.

Move through each section in order, without overthinking any single step:

Section A — Doodle: a fast, reflexive first mark. Don't plan it, just let your hand move.
Section B — What is it?: tap any subject tags that feel right (structure, water, movement, etc.) and add free text.
Section C — Sensory data: pick from the dropdowns for colour, sound, texture, temperature, and so on — whatever comes, not what you'd expect.
Open Collector: a stream-of-consciousness box — write anything, don't edit or judge it.
Defined Collector: list specific discrete signals one at a time (S1, S2, S3...) — you can also upload a photo of a paper sketch here if you worked part of the session offline.
Matrix: a structured table for tighter shape/energy/texture/etc. breakdowns per signal.
Section D / Final Sketch: drawing tools with color and size options, plus an upload button if you'd rather bring in a scanned sketch.
Summary: pull it together — a short written impression of the target as a whole.

When you're done, hit the large 👁 Reveal button at the bottom to see the actual target. Rate the session — Miss / Partial / Hit / Strong hit — then 💾 Save to log. You can also 🧹 Clear session at any point to wipe everything — doodles, tags, text, and the active target — or hit ↺ New to reset and pull a fresh target.

The 🎯 Targets tab holds your whole blind pool — this view breaks the blind, so only open it between sessions, not mid-session. From here you can add a new target (label, type, coordinates, URL, image, notes — only Label is required), edit any existing one in place, import/export CSV for bulk editing in a spreadsheet, 🔄 rotate every target's blind code at once while keeping the same images (useful for keeping a pool blind after repeated use), or 🗑 delete all to start fresh.

Every saved session shows up in the Log tab with its doodles, sketches, sensory data, and summary all in one place. You can re-rate any entry at any time using the small Miss/Partial/Hit/Strong hit buttons under its date — handy if you want to revisit a call. 📄 Export PDF gives you a printable/archivable copy of the whole log.

The Charts tab shows your hit rate over time (with a rolling average trend line), a breakdown by target type, and an overall score distribution — once you've logged at least 3 sessions.

Inside a session, there are two buttons — 🔮 AI guess what the target is (before reveal) and 🤖 Get AI match analysis (after reveal) — that send everything you've filled in so far to Claude and get back its own independent read, so you can compare your decode to the AI's.

These need your own Anthropic API key to work — a Claude.ai subscription (Pro, Max, etc.) does not include this; API access is billed separately, per token, through a different account. For light use (a guess + feedback call per session, a few dozen sessions a week) the cost should be small — likely just a few dollars a month — but it's a real, separate charge once you set it up.

To get a key, when you're ready:

1. Go to console.anthropic.com and sign up (or log in).
2. Verify your email and fill in the basic profile questions.
3. Add a payment method under Billing — the API has no ongoing free tier.
4. Go to Settings → API Keys → Create Key, name it something like "RV app," and click Create.
5. Copy the key immediately — it's a string starting with sk-ant-, and it's only shown once.

Your key is never written into this app's files or pushed to GitHub — it's stored only in this browser's local storage, on this device, and sent directly from your browser to Anthropic when you use the AI buttons. It's the same key used by the ARV app, so you only need to save it once. If you ever want to remove it, use the Clear button below.

The 🤖 AI Viewer tab lets an AI remote viewer (by default called Vesper) run a blind session, one stage at a time, while you act as the monitor. The AI is only ever sent the target number. The target's name, link, notes and photo stay in your browser until you reveal.

Before your first session: settings

1. Save an API key in the card above. Claude models need the Anthropic key; GPT models need the OpenAI key.
2. Open the 🤖 AI Viewer tab and press ⚙️ Settings (top right of the viewer's card).
3. Name: what the AI viewer is called. Your name: what it calls you (optional).
4. Model: which AI runs the sessions, the chat and the guess. Claude Sonnet is a good default. Opus is slower and costs more; Haiku is fastest and cheapest. Under ChatGPT you can pick a GPT model instead. A line at the bottom of Settings tells you whether the right key is saved for the model you picked.
5. Personality & background: who the AI is, written as "You are…". Press ↺ Default personality to restore the original.
6. Protocol reference (Farsight SRV): the written SRV steps the AI follows in every session. Edit it freely, or press ↺ Default reference.
Settings save automatically as you type.

1 · Choose the target

🔀 Random Target from My Targets list picks a blind target from your 🎯 Targets pool. ✍️ Target Not in the Targets List is for a target someone else tasked: type the target number (or press 🎲 to make one). Name, URL, type and notes are optional and stay hidden until reveal; tick Also save to my random target list to keep it in your pool. Then press ▶ Start session.

2 · Run the session, one step per click

Each step shows the exact prompt that will be sent. Edit it if you like (↺ Original wording restores it), then press ▶ Ask. The steps are: Welcome, Stage 1 · Initial contact (with ideogram sketch), Stage 2 · Sensory data, Stage 3 · Dimensions & sketch, Stage 4 · Structured matrix, Stage 5 · Qualitative probing, the optional Perspectives A and B (press Skip this step to leave them out), and Lock session, which compiles the record. Earlier notes are never changed. The progress chips under the target code show where you are; ⏸ Stop cancels a slow answer.

Between steps you can send an extra monitor prompt: tap one of the neutral suggestions (e.g. "Describe any structures.") or type your own. Keep it neutral, never leading. It is saved as a supplementary entry.

3 · Guess, reveal, judge and score

• 🔮 Ask the AI to guess the target (before reveal): it reads its own locked record and names what it thinks the target is, with key signals, an alternative and its confidence. The guess is saved with the session.
• 👁 Reveal shows the target. For an outside target, fill in what it was so the judge has something to compare.
• ⚖️ Run blind judge: a separate, fresh AI that took no part in the session compares the raw stages with the target and its photo, and lists matches, partial matches and misses with a score. The viewer never grades its own session.
• Add your own rating (Miss / Partial / Hit / Strong hit) and save the session to the 📓 Log.
• Don't know the target yet? Press 💾 Save now, reveal later. Later, open the session in the 📓 Log, press I know the target now, then run the blind judge there. You can also ask for the 🔮 guess from the Log while it is still unrevealed.

Talk with the AI viewer

At the bottom of the tab you can chat outside of sessions. It remembers the conversation and has a practice journal of its past sessions (their ratings and judge scores), the full stage records of its 3 most recent sessions and of any session whose number you mention, and the session in progress. It is never told what the targets were. Clear conversation empties the chat; 🗑 Clear memory in Settings also starts the journal fresh (your saved sessions stay in the Log).

Track record

The track record card counts the AI's sessions, the share you rated Hit or better, and the share the blind judge scored Hit or better.

Every step is a separate API call billed to your key, so a full session with guess and judge costs a little more than the 🔮 and 🤖 buttons on the Session tab.

Nothing here requires a manual "Save" — every target and session is written to your browser's local storage the moment you add or log it, and it'll still be there next time you open this app on this device. Local storage lives only in this browser, so if you want a portable backup, use ⬇ Export all on the Targets tab or 📄 Export PDF on the Log tab.

Select a session below to score it.

Pick any saved session to load or update its score card.