Know what to askbefore you spend.
8 research methods, each borrowed from someone who published theirs and wired to these tools by name: which call to make, in what order, and the number that decides the answer before the data arrives.
Or hand them to your agent
$ curl 'https://wutheringai.com/v1/free/skills'Or install them in Claude Code
$ /plugin marketplace add WutheringAI/claude-pluginInstalls every method as a skill you can call by name, plus two more that spend nothing.
One run · cheapest killer first
- 1 · Harvest the language2 calls
- 2 · Price the demand2 calls
- 3 · Find who already pays3 calls
- 4 · Read the buyer’s words2 calls
- 5 · Size what is left3 calls
Kill criteria — fixed before the data arrives
The deliverable
A verdict, and the number that decided it.
8
methods, one job each
$0
to read one, with no account
~$2.60
in calls for a typical run
01 / pick your job
What are you about to decide?
Each one names its own kill criteria, so a run can end early with a real finding instead of late with a hedge — and states what it spends before you start it.
Is there real demand for this?
“I have an idea. Tell me whether anyone is already paying to solve this, before I spend a quarter building it.”
Gives you A verdict — validated, refuted, or inconclusive — the single number that decides it, the two strongest pieces of evidence for, the strongest against, and the cheapest next call that would change your mind.
Read it — freeWhere is the unoccupied position in this category?
“We know the category and we know the incumbents. Map the forks this category has already run, show me who took which branch and which branch killed them, and tell me which one is still open — with the price, the bundle and the message that go with it.”
Gives you The maze as a tree: the forks this category has already run, who took each branch, which branches are dead and which walls have since moved — plus the one open position, scored on demand, intent and occupancy, and the offer that takes it: buyer, bundle, price metric and message, with what it deliberately gives up and how long it stays yours.
Read it — freeWho are we really up against?
“Prospects do not understand why they would pick us. Tell me what they would do if we did not exist, and what makes us obviously the right choice for the ones who should buy.”
Gives you A filled positioning statement — alternatives, unique attributes, value, best-fit customer, category — plus the competitor's tested claims, their untested ones, and the silence you can own.
Read it — freeWhat do we charge, and does the money work?
“Tell me what to charge, what to charge for, and whether the unit economics survive a conservative case.”
Gives you A price, a value metric, a packaging split into leaders, fillers and killers, and the LTV:CAC figure in three scenarios with the conservative one shown first.
Read it — freeWhere do the next 1,000 customers come from?
“Pick one acquisition channel we can afford, prove it works for a product like ours, and tell me its ceiling before we commit a year to it.”
Gives you One named channel with its CAC, its reachable ceiling, the product change it requires, and the measured evidence that it works for a product with these characteristics.
Read it — freeWhat should we build next?
“Tell me what to build next from what users actually say when they switch, not from what they say in a survey or what we wish they said.”
Gives you Themes ranked by size, each carrying its force, three verbatim quotes, a measured demand figure, and the specific change that would move it — split 50/50 between deepening what works and removing what blocks.
Read it — freeIs this company actually growing?
“I am about to bet on this company — invest in it, partner with it, sign a contract with it, or take the offer. Tell me from the outside whether it is actually growing, which engine is carrying it, and what I would only learn after I signed.”
Gives you A trajectory verdict with the series under it: branded demand year over year against the rival set, organic value and footprint over 24 months, which engine carries the growth and whether it was earned or bought, what payroll says ships next, and the three questions only a data room can answer.
Read it — freeWhat would prove us wrong?
“The plan rests on one belief nobody has tested. Tell me the observation that would prove it wrong, then go and look — and tell me whether the refutation is already sitting in public data.”
Gives you A verdict of refuted, survived or untestable — with the belief as it was written before the first call, the rival explanations it was tested against, the observation that discriminated between them and what it returned, the variation you were tempted to make and did not, and the one prediction a surviving explanation makes that you have not yet checked.
Read it — freeEvery skill is plain text over a plain GET. Add ?format=markdown and it comes back as an agent skill file, frontmatter first; add the skill’s own inputs — ?idea=… — and it comes back already aimed at your job.
02 / one finished run
What $2.60 of calls actually buys.
A real run of the demand-validation skill, made against this API on 30 August 2026 and reproduced without editing.
The brief
An AI scribe that writes clinical progress notes for private-practice therapists.
skill=demand-validation · market=United States · run 30 August 2026
The verdict
Not validated.
The demand is real and someone is already being paid for it. What fails is the channel the data prices: at a $50 blended CPC against a $40/mo price, paid search returns 37¢ on the dollar — and 86¢ even if everything goes right.
LTV : CAC — measured against the threshold fixed before the run
15
calls attempted
13
calls charged
2
failed, refunded
$2.60
total spent
How it got there
In the skill’s order — cheapest killer first, so a run that is going to fail fails before it spends. Two calls returned nothing and were refunded; both are still here.
- 1Harvest105 candidate terms, none of them invented
keyword_ideaskeyword_suggestions$0.40 - 2Price3,340/mo of commercial intent, at a $50 blended CPC
keyword_volumekeyword_intent$0.40 - 3Money3 advertisers paying — one for 606 days straight
ads_advertisersads_creatives$0.80 - 4SalaryNothing: the listing had expired. Refunded.
company_jobsrefunded - 5ListenA $40/mo ceiling, said out loud, unprompted
web_searchsocial_reddit_comments$0.40 - 6TrendThe AI-specific queries are shrinking, not growing
keyword_history$0.20 - 7SizeA competitor that read as dead was on another TLD
domain_traffic$0.40
What the market is worth
Measured by keyword_volume and classified by keyword_intent, in one call each.
searches / month · US
- ehr for therapists720$100 cpc
- emr for therapists720$100 cpc
- practice management software for therapists590$75 cpc
- ai therapy notes260$47 cpc
- therapy notes software210$42 cpc
- hipaa compliant ai scribe210$28 cpc
- soap note generator210$24 cpc
- ai scribe for therapists140$68 cpc
Eight of forty terms shown. The commercial and transactional cluster totals 3,340/mo once free-template seekers are excluded — above the ~1,000/mo bar this skill sets for a B2B tool, and nowhere near the ~5,000 it sets for consumer.
Who is already paying
The top rung of the evidence ladder: nobody sustains a losing ad for a year. From ads_advertisers and ads_creatives.
61
creatives from one competitor
49
still running after 90 days
17
running over a year
606
days — the longest single run
And the evidence against
Stated before the conclusion, because the skill requires it. Twelve months of keyword_history: the AI-specific queries are shrinking while the spend is not.
- ai scribe for therapists720 → 50-93%
- ai therapy notes390 → 170-56%
- hipaa compliant ai scribe170 → 140-18%
- therapy notes software210 → 140-33%
Aug 2025 → Jul 2026, US. The demand did not disappear: it moved into brand terms, which two of the incumbents own outright.
In the buyer’s own words
One call to web_search to find the thread, one to social_reddit_comments to read it. Each quote is coded to the force it represents.
“I’m in solo practice and totally get the note fatigue.”
“Mentalyc — bit cheaper but messy interface, and it takes 10–20 minutes to upload and generate a SINGLE note — basically unuseable for me.”
“Some AI notes platforms freaking charge EXTRA for anything outside of SOAP and DAP formats, which I think is bullshit b/c they’re not better there just different.”
“I just copy the notes into SimplePractice.”
“There are several that are over $40/month which are too expensive for me to even try, so I left those out.”
Against the bar, set before the data arrived
- Reach3,340/mo of commercial and transactional intent, against a B2B bar of ~1,000. Narrow, but it clears.
- TriangulationFour independent families agree that buyers exist: search volume, sustained ad spend, eight named products, and unprompted complaints.
- ExplanationNote fatigue in solo practice, a hard $40/mo ceiling, and HIPAA as the gate. The story survived the data.
- ForcesPush and pull are well evidenced. Habit — the notes already live in an existing practice system — is named in the quotes and never sized.
- EconomicsA blended $50 CPC at a 2% funnel is a $2,500 CAC against a $914 LTV. LTV:CAC 0.37×, and 0.86× even in the optimistic case. The bar is 3.0×.
The cheapest next call
Find out what the 606-day advertiser is actually bidding on, and at what contract value. That is the channel-bet skill, and it is the next $3 well spent.
Where the instrument was blind
One competitor came back with zero organic keywords — which reads like a company with no search presence and was in fact the wrong top-level domain; the correct one had 4,858 keywords and $37,542 of monthly traffic value. The salary step never ran: the listing had expired, so the call failed and was refunded. Both are in the report because a report that hides its gaps cannot be checked.
03 / what a run costs
Priced before you start, not after.
Reading the skill is free. Only the calls it tells you to make are charged, and only when they succeed.
- Is there real demand for this?
demand-validation13 calls · $2.60stop at $4.80 - Where is the unoccupied position in this category?
idea-maze18 calls · $3.60stop at $6.20 - Who are we really up against?
positioning-teardown15 calls · $3stop at $5.40 - What do we charge, and does the money work?
pricing-power12 calls · $2.40stop at $4 - Where do the next 1,000 customers come from?
channel-bet13 calls · $2.60stop at $4.80 - What should we build next?
switch-evidence11 calls · $2.20stop at $4 - Is this company actually growing?
company-diligence13 calls · $2.60stop at $4.40 - What would prove us wrong?
crucial-test9 calls · $1.80stop at $3.20
Figures are the published per-call price multiplied by the skill’s own call budget. Failed calls are never charged — the run above attempted 15 and paid for 13.
Hand one to your agentand let it work.
Reading them is free and needs no account. The calls they tell you to make start with $5 of credit, on us — enough for several runs before you decide anything.