The hypothesis problem
You have opinions. You have no versioned edge.
"SOL longs in Asia when funding negative", you said it in Discord six weeks ago. You cannot prove it worked. You never saved it. You sized from memory yesterday and it was not the same rule.
That is not a intelligence problem. That is a hypothesis problem.
Edge is not a tweet. It is a versioned, trackable claim with discovery stats and live scorecard. Workbench discovers it. Edge rules store it. Hypothesis inbox promotes it. Session Desk runs it.
Without versioning, every week you restart from opinions. With it, research compounds.
Pro is $99/month. Paid from day one.. Or $99/mo if you are ready. Cancel in one click from settings.
- Edge rules + inbox
- Workbench Alpha Loop
- Live scorecard
Opinions rot. Edge rules compound.
You need versioned hypotheses, not another hot take.
What most traders run
- Twitter / DiscordThesis with no sample size
- One-off backtestCherry-picked window
- Memory"I think funding works on ETH"
- Live bookNo tag linking trades to thesis
Feels like edge. Cannot survive contact with your actual journal.
What Thrive tracks
- DiscoveryWorkbench save_edge_rule + stats
- VersioningHistory when rules change
- Live scorecardJournal tags vs backtest
- PromotionHypothesis inbox → Session queue
Same login as signals and journal. Trader context injects rules into coach and Session Desk free.
A backtest without live tracking is a story. A live book without discovery stats is noise. Edge requires both windows connected.
When a rule decays, live win rate drops vs discovery, nothing in a spreadsheet alerts you. Edge scorecard and nightly digest flag decay.
Alpha Loop in Workbench validates patterns with sample size, profit factor, and win-rate lift before save_edge_rule.
Validated hypotheses land in inbox for Session Desk. Deactivate rules that degrade, versioning preserves what you learned when you change your mind.
A month of opinions with no scorecard
Week one: you discover funding mean-reversion on alts works in your manual review. Week two: you trade it twice, mixed results, no tags. Week three: you argue about it in Telegram with zero linked trades. Week four: you abandon the idea and chase a new narrative.
If the rule had been saved with backtest stats, journal strategy tags, and live scorecard, you would know by week two whether your execution matched discovery, or whether the edge was fiction.
That is the hypothesis problem. Research without versioning is reset every month.
Which hypothesis gap sounds like you?
Pick the one that stings. That is where edge rules pay first.
The thesis trader
Strong narratives, weak receipts. save_edge_rule forces sample size and stats before a hypothesis becomes a rule.
Start here: Pro+ Workbench
The one-off backtester
You ran SQL once, got excited, traded live without live scorecard. Edge rules connect discovery window to journal tags.
Start here: Pro + Workbench SQL
The no-scorecard pro
You journal but never tag strategy to rules. Performance edge scorecard shows live vs backtest, decay visible early.
Start here: Pro journal + edge rules
Why opinions feel like edge until they are not
Narrative confidence is not sample size. Workbench Alpha Loop requires validated patterns before promotion. Hypothesis validate jobs run async when you need rigor.
Journal tags without rules are orphaned data. Strategy labels link to edge rules for scorecard. Otherwise you cannot roll up performance by thesis.
Session queue without inbox repeats yesterday's scroll. Hypothesis inbox surfaces validated edges for Session Desk, research picks up where it left off.
The fix is not more opinions. It is versioned edge rules with live tracking and inbox promotion.
Reproduce Analysis SQL (credits apply) reruns saved snapshots for trust and provenance, same query, same conclusion, weeks later.
You have "proven" setups before
Every trader has a setup that worked three times and became identity. Without live scorecard, you keep trading it through decay because memory only keeps wins.
Edge rule version history shows when you changed conditions, and whether the change helped. That is how research compounds instead of mythologizing.
Trader context injection puts active rules into Session Desk, Trade Builder, and coach prompts at zero marginal credits, your edge follows you.
What versioned edge actually shows
Not "trust me bro." Discovery win rate, profit factor, sample size, then live stats on tagged trades.
Decay alerts when live underperforms discovery by ten points or more.
- Backtest stats vs live scorecard on same rule
- Hypothesis inbox: proposed vs validated promotion funnel
- Edge rules flagged for rolling win-rate decay
- Session queue aligned to active rules only
Pro+ Workbench unlocks full Alpha Loop, deep research, and complex backtests.
Thrive versions the edge
Discover → save → track → promote → session → debrief.
Discover
Workbench SQL + Alpha Loop. save_edge_rule after validation.
Track
Edge scorecard + journal strategy tags. Version history on changes.
Run
Hypothesis inbox → Session Desk queue. Coach reads the same rules.
Signals tied to hypotheses
Context and provenance, not isolated alerts.
Journal tags linked to rules
Live scorecard needs tagged trades, seconds to tag.
Your first 30 days versioning edge
One rule
Save one edge rule manually or from Workbench. Tag every trade to it. Baseline live stats.
Validate or kill
Run hypothesis validate if needed. Promote validated to inbox. Deactivate if live diverges from discovery.
Compound
Second rule from Alpha Loop. Session queue from inbox. Coach on Pro+ references your rule book.
Who this is for
Traders who generate ideas faster than they track them. If you never backtest or tag, start with Loop Problem and Session Problem first. If you backtest but never live-track, this page is yours.
Pro+ is the natural fit for Workbench depth; Pro can run manual edge rules and journal tags.
One month of untracked opinions vs Pro+
Trading a decayed "edge" for one month often costs more than a year of Workbench access.
Credits fund agent depth, Reproduce Analysis, deep research, not platform-funded regime overlays and trader context injection.
What you get in Pro
- AI review of your own trades on the top 20 assets
- Trade journal with emotion and strategy tags
- CSV import and read-only account sync
- Performance dashboard and equity curve
- Up to 50 custom alerts
- Macro event calendar
More in the Cursor for Trading series
Cursor for Trading: The Problem Series: pain-led guides to the closed research loop, Session Desk, versioned edge rules, and why terminals are not the answer.
Your Charts Show the Market. Nothing Shows You.
Charts show the market. Nothing shows your book. Journal fills next to context and see which habits pay. Pro is $99/mo. Pro+ is $199/mo today.
You Research Everywhere. You Close Nowhere.
Research, data, and the fill should live in one loop. Journal the close so the next size has a record. Pro is $99/mo. Pro+ is $199/mo today.
You Have Infinite Intel. You Have No Session.
Charts and signals without a session are noise. Brief, queue, plan, then review the book. Pro is $99/mo. Pro+ is $199/mo today. Cancel anytime.
You Do Not Need Another Terminal.
You do not need another terminal. You need the fill next to the reason. Journal and review in one place. Pro is $99/mo. Pro+ is $199/mo today.
Product docs: Session Desk · Edge Rules · Pricing
Questions before you version edge
You do not need another opinion. You need a versioned edge.
Explore Pro+ at $199/mo for Workbench depth, or Pro at $99/mo, paid. There is no free trial. Your last untracked thesis probably cost more than a month of versioning.
Pro is $29/mo for your first 3 months, then $99/mo. Paid from day one. Go to checkout
Risk disclosure. Thrive is a market intelligence and trading analytics platform. We do not provide financial, investment, or trading advice, and we do not place trades or hold customer funds. Trading cryptocurrency involves substantial risk of loss and is not suitable for every investor. Past market behavior does not guarantee future results. Nothing on this page constitutes a recommendation to buy or sell any asset.