idea-001 · human · US
The device is indicated for the treatment of stress urinary incontinence in cognitively intact adult women in the home setting, by prescription, following a documented failed four-week trial of pelvic muscle exercise training.
Finished enough to read, and waiting for you. 7 of 7 departments answered: 5 conditional, 2 abstaining, none blocking. 27 of 218 claims verified, 4 refuted. 4 of 8 things it rests on are critical and open; the first is cheap to find out.
A prescription home device for women with stress urinary incontinence who have already tried pelvic floor exercises and not got better: a small probe used internally that stimulates the pelvic floor muscles and reads their electrical activity, so that stimulation is triggered by the woman's own contraction. That much is no longer new: a prescription device with EMG (electromyography)-triggered stimulation through a vaginal probe is already sold in the US claim 167refuted, and cleared devices in the same FDA (US Food and Drug Administration) product code already carry the mode claim 153verified, claim 154verified. So the opportunity changed shape today. The P2 the discriminating trigger device adds a skin electrode the woman places on her abdomen each session, so that a burst is withheld when the detector sees her straining (a Valsalva) or bracing rather than squeezing the right muscles. Whether that difference is real is the whole question: whether no sold device already does it claim 173unconfirmed, whether a self-placed two-channel detector can do it claim 204unconfirmed, and whether a prescriber would choose it claim 174unconfirmed. The money is unchanged by the pivot: Medicare pays this device class as a monthly rental, $74.52, the same for any device under the code claim 193verified, so the difference earns no higher price — only, if anything, a larger share — while adding an electrode every session.
A urogynecologist or pelvic floor physical therapist prescribes a woman with stress urinary incontinence a home stimulator after she has failed four weeks of pelvic muscle exercise. A DME (durable medical equipment) supplier rents it to her under E0740 (a HCPCS code for durable medical equipment), and Medicare or a commercial plan pays the supplier a fixed monthly allowable claim 9verified, claim 193verified. The prescriber picks it instead of a cleared single-channel EMG-triggered device: the MyOnyx is listed and offered for US sale claim 167refuted, K243079 (an FDA 510(k) clearance number) is cleared for prescription home use claim 153verified, and K220161 is cleared over the counter. The reason to pick it is that those triggers fire on the pelvic floor channel alone claim 202verified. Pelvic floor EMG also rises on a strain claim 150unconfirmed, so they may stimulate the manoeuvre the therapy exists to correct. This device withholds stimulation when the abdominal reference says she is straining. The shape is chosen because it is the only answer on the record to the specific failure IR-56 claim 167 came back refuted found: the closed loop is sold already, and a differentiator has to name something the sold implementation does not do. The bet is that the withhold is real in these women, that she can run it alone at home, that no incumbent already does it, and that the rental money clears the cost of an extra electrode every session.
In the Wrap's one-paragraph version, written for someone who has not opened the file. Domain: stress-urinary-incontinence
The budget-holder chain, end to end. The manufacturer does not bill a payer. A DME supplier does.
What this idea rests on that is not yet known, in the order this idea's own decision queue puts it: worst if false first, cheapest to settle first inside that. The first row is open; open any other for what would settle it.
What would settle it (cheap): REQ-0194 Run the twenty structured prescriber calls…, the twenty structured calls this line has named since 2026-09-14 and which had no request behind them until today, preceded by REQ-0062 Read the United States guideline corpus for…'s read of ACOG (American College of Obstetricians and Gynecologists), AUGS (American Urogynecologic Society) and AUA (American Urological Association)/SUFU for whether any US guideline names electrical stimulation
What would settle it (project): REQ-0177 For the P2 device as prescribed at home… (sessions per course), then REQ-0176 Under P2 (the landed monthly electrode cost against the $74.52 rental), with REQ-0173 Retrieve deciding whether the supplier carries it inside the rental. Then the denominator this file has never had: REQ-0046 Build the market model this candidate has…'s market model and REQ-0060 When the market model REQ-0046 commissions…'s build volume by object, and the OIG (Office of Inspector General) supplier-acquisition-cost reports for claim 122unconfirmed's ratio
claim 88unconfirmed, claim 48unconfirmed, claim 111unconfirmed, claim 67unconfirmed, claim 57unconfirmed, claim 121unconfirmed, claim 87unconfirmed, claim 122unconfirmed, claim 141unconfirmed, claim 113unconfirmed, claim 117unconfirmed, claim 139unconfirmed, claim 181unconfirmed, claim 183unconfirmed, claim 208unconfirmed, claim 14unconfirmed, claim 92unconfirmed, claim 108unconfirmed, claim 116unconfirmed
What would settle it (project): REQ-0050 Reconcile the adherence evidence for…, the published discontinuation and withdrawal figures read as one series, then a twelve-week continuation arm in the first clinical work, which now has to include placing the reference every session. REQ-0160 For app-connected home pelvic floor devices adds capture rates on app-connected home pelvic floor devices. None of it can be answered without patients
claim 26unconfirmed, claim 27unconfirmed, claim 28unconfirmed, claim 59unconfirmed, claim 80unconfirmed, claim 138unconfirmed, claim 169unconfirmed
What would settle it (project): counsel's read under REQ-0048 Do the issued claims of US10307589 (accepted by Legal & IP (intellectual property)) of the four grants' independent claims against a pre-burst design. It uses the claim text REQ-0191 From USPTO public records retrieves from USPTO (United States Patent and Trademark Office), the expired alternating disclosure (REQ-0053 US6240315 and US6289245, REQ-0153 Read the full specifications of US6240315…), and K053434's possible 2006 ETS clearance (IR-75 claim 201). Then REQ-0185 A US landscape scoped to pivot P2's new…'s landscape for the withhold element
What would settle it (project): literature first: REQ-0055 Search for any reported sensitivity or… and REQ-0187 Search for any reported sensitivity for any published single- or two-channel discrimination figure, REQ-0170 Search the literature for whether the wrong manoeuvre survives the failed exercise trial, REQ-0193 Search PubMed for any outcome comparison, and REQ-0181 Is an intravaginal NMES burst delivered… for whether a burst on a strain is a harm (claim 209unverified). Then a sensing-only recording in 10–15 women on a purchased two-channel unit against the pass line REQ-0189 State the acceptance criterion the P2… asks Clinical to set, and REQ-0197 Specify's formative placement study on a works-like prototype. The outcome half (claim 178unconfirmed) has no test short of a controlled comparison against single-channel triggering, which is funded and has no sponsor
claim 150unconfirmed, claim 105unconfirmed, claim 176unconfirmed, claim 204unconfirmed, claim 172unconfirmed, claim 100unconfirmed, claim 207unconfirmed, claim 209unverified, claim 178unconfirmed, claim 179unconfirmed, claim 33unconfirmed, claim 91unconfirmed, claim 96unconfirmed, claim 109unconfirmed, claim 101unconfirmed, claim 103unconfirmed, claim 147unconfirmed, claim 149unconfirmed, claim 114unconfirmed, claim 145unconfirmed, claim 146unconfirmed, claim 142unconfirmed, claim 144unconfirmed, claim 102unconfirmed
What would settle it (cheap): REQ-0166 Read the manufacturer labelling item (a) and REQ-0190 Read device labelling no connector reaches… item (a), the four products' labelling for how each trigger fires, which IR-58 claim 173 decided is the last commission on claim 173unconfirmed. REQ-0192 Read in full reads the two Chinese disclosures that could refute claim 199unconfirmed. REQ-0194 Run the twenty structured prescriber calls…'s twenty prescriber calls ask claims 168 and 174, and REQ-0188 Candidate claims 187 and 202 place the… asks Regulatory whether an incumbent's firmware change needs a new 510(k) (premarket notification)
claim 173unconfirmed, claim 202verified, claim 160unconfirmed, claim 199unconfirmed, claim 174unconfirmed, claim 168unconfirmed, claim 110unconfirmed
What would settle it (cheap): REQ-0172 Retrieve (the PDAC (Pricing, Data Analysis and Coding contractor) E0740 list, including which components each listed system names) and REQ-0195 Call DME suppliers that dispense home… (DME supplier calls). Then REQ-0020 Retrieve's remaining commercial policies and REQ-0154 From CMS primary sources's Medicare Advantage share. The coding determination itself needs a PDAC application on the built, cleared device
claim 9verified, claim 55unconfirmed, claim 180unconfirmed, claim 56unconfirmed, claim 170unconfirmed, claim 90unconfirmed, claim 32unconfirmed, claim 86unconfirmed, claim 89unconfirmed, claim 74verified, claim 75verified, claim 129unconfirmed, claim 58verified, claim 193verified, claim 194unconfirmed, claim 182unconfirmed, claim 77verified
What would settle it (project): REQ-0196 Retrieve from fda.gov and quote (the 510(k) Program guidance's decision points), REQ-0179 May the P2 abdominal reference electrode be… (whether the electrode is a third-party cleared part or sits inside this 510(k)), REQ-0183 Retrieve from fda.gov and quote (critical task) and REQ-0198 Retrieve and quote (the isolation clauses). Then an FDA pre-submission written response, the only thing that settles claims 155 and 188
claim 2verified, claim 6unconfirmed, claim 153verified, claim 154verified, claim 155unconfirmed, claim 188unconfirmed, claim 187unconfirmed, claim 189unconfirmed, claim 190unconfirmed, claim 184verified, claim 206unconfirmed, claim 192unconfirmed, claim 205unconfirmed, claim 156verified, claim 158verified, claim 165verified, claim 97unconfirmed, claim 36unconfirmed, claim 115unconfirmed, claim 161unconfirmed, claim 162unconfirmed, claim 164unconfirmed
Each row is one thing this idea has to be true. Ordered by what a failure would cost and, inside that, by what it costs to find out — for this idea only; bands never compare between ideas.
Seven chairs, each answering from its own field. The verdict, and the first condition it set.
| Regulatory | conditional | C1 A device with this intended use is eligible for 510(k) and needs neither De Novo (De Novo classification request) nor PMA (premarket approval). — claim: 6 — would be settled by: a pre-submission written response. At P2 the discriminating trigger the pre-sub has to show the… +20 more condition(s) the position → |
| Clinical & Scientific | conditional | C1 No randomised comparison of gated against open-loop intravaginal stimulation exists in SUI (stress urinary incontinence), so the margin in claim 33unconfirmed has never been measured. Under P2 the discriminating trigger this is a necessary condition… +16 more condition(s) the position → |
| Market Access | conditional | C1 Met, and retired 2026-09-14. NCD 230.8 covers this indication — claim: 9 — would be settled by: full-text retrieval of §230.8. That was done; the tag is the Verifier's. Kept as the record of what… +9 more condition(s) the position → |
| Technology & R&D | conditional | C1 A threshold-crossing trigger needs EMG sensed before the burst, not during it. — claim: 100 — would be settled by: the existence of devices built that way. K201290's summary now describes it ("As… +23 more condition(s) the position → |
| Operations & NPI | abstains | C1 The candidate is at least four manufactured or delivered objects with different lifetimes, so a single per-unit cost of goods as claim 48unconfirmed is worded cannot be stated without naming the… +13 more condition(s) the position → |
| Legal & IP | conditional | C1 No live, enforceable US patent reads on intravaginal NMES (neuromuscular electrical stimulation) triggered by pelvic floor surface EMG in a way that blocks this design. P2 the discriminating trigger's reference channel does not change this, because it is additive.… +8 more condition(s) the position → |
| Commercial | abstains | C1 The payer-mix question is asked against a three-way denominator: Medicare fee-for-service, Medicare Advantage and commercial. MA is Medicare-primary, is outside the FFS (fee-for-service) book claim 70verified is… +6 more condition(s) the position → |
A position is not evidence: it cites the file's tags and issues none. blocks opens an impact review and kills nothing.
Only verified counts toward a score. A low count is usually a statement about the tooling — an absent data file, an unset key — rather than about the idea.
4 claim(s) the Verifier found to be false — a refuted claim is a finding, and the loop owes each one an answer.
there are two, they answer different questions, and neither is rewritten to match the other.
Not a test of either kind but cheaper than both, and the one the viability deck rests on: a read of the MyOnyx and NeuroTrac user manuals for whether their trigger already uses a reference channel (REQ-0166 Read the manufacturer labelling(a), REQ-0190 Read device labelling no connector reaches…(a)). IR-58 claim 173 has committed that if one does, the next review is a pivot review on differentiation.
Where the evidence actually stands (2026-10-01, from wrap_status.py). 27 of 218 claims verified — 177 unconfirmed, 10 unverified, 4 refuted. Of the 27 verified, 6 are registry facts, 3 are published findings, and 18 carry no evidence-class stamp. "9/9 factors scored · 4/4 core" counts the Stage 2 triage table and is not the verified figure. The Verified Composite, recomputed for P2 the discriminating trigger, reads 53/100 over 4/9 factors verified (Regulatory 4/5, Reimbursement 3/5, TAM 2/5, Moat 1/5). It fell from 60 only because Moat entered the average: what P2 the discriminating trigger adds over the cleared triggered devices is detector logic plus a commodity electrode claim 202verified, claim 184verified. At 4/9 the candidate is below the 6/9 that Stage 5 would gate on. TAM still sits exactly on its 2/5 kill floor. The ## Gate Check section above predates today's numbers.
The ten unverified claims (209–218) were appended today by the assessment, the decks and the business case and have not had a Verifier pass.
Pivot options on the table: Shapes this opportunity could take and has not taken. Nothing here is chosen. A pivot is the outcome of an impact review or a recorded redesign, and the front end does not select a shape over a chair — least of all over…
Recorded under P1 commercially-insured first on 2026-09-14 with a named owner and unchanged since. Commercial's abstention on 2026-09-19 did not move it: the chair named the absence of a market model as the one thing that would move it off abstains, and that is REQ-0046 Build the market model this candidate has…. Market Access has declined to argue for it in writing ("the price is the problem and the price is Commercial's call").
Contributed by Clinical & Scientific on 2026-09-19 in its ## If the opportunity pivots, and explicitly not chosen by that chair either: "It is not my call to make and I am not making it; it is the pivot option this department contributes, with the evidence for it named."
The shape. The device is positioned as the adherence intervention for guideline-standard supervised pelvic floor muscle training — twelve weeks rather than four — with adherence at twelve weeks as the primary endpoint and symptom change secondary, instead of as a therapy that rescues a woman in whom PFMT (pelvic floor muscle training) has failed.
The evidence the chair named for it, as that chair's reading and not as a finding of this file. Every line here is a claim in the candidate, and every one of them is unverified or unconfirmed; they are referenced in words because a tagged citation in this record declares a dependency, and an option is not a dependency:
What choosing it would cost, recorded so the option is not cheaper on this page than it is in life. A different intended_use:, which is the one frontmatter field a pivot may change. A different primary endpoint and therefore a different clinical plan. A re-statement from Regulatory, because an adherence indication is a different predicate conversation and this file's own record is that the same hardware needed a second 510(k) to change one word of an indication; and from Market Access, because the four-week precondition exists to mirror NCD 230.8's eligibility language and an adherence device is not obviously the thing that language describes. It would also change what the closed loop is for: gating exists to make stimulation contingent on a voluntary contraction, and an adherence device is judged on whether a woman keeps using it rather than on whether the stimulus was gated. Whether that makes the differentiator more defensible or redundant is Technology & R&D's and Legal & IP's to say, and Technology has never stated.
Technology & R&D answered that on 2026-09-19 and the answer has two halves that only mean something together, so they are recorded together. The chair's words are "redundant, and cheaper for it." First half: every hard engineering problem in its position exists only because stimulation is made contingent on a correctly performed voluntary contraction — the trigger-specificity question at A34 No published study reports the sensitivity…, the supervision gap at A38 The only published randomised trial of…, the timing budget at A30 The therapeutic stimulation envelope… and A31 Real-time recovery of volitional surface…. Remove the contingency and what is left is an instrumented open-loop stimulator with a session log, which is claim 35unconfirmed's sixteen days in thirty and nothing harder, and A37 Automatic capture of at least 16 days of… says that constraint is behavioural rather than technical. Second half, and the chair insisted the two be read as one sentence: a device with no gating is the KPI predicates with a phone app. It strands the differentiator entirely. So Option 2 the adherence device rather than the… is "the cheapest shape to build and the hardest to distinguish", and a reader who takes only the first half away has taken the half that makes it look free. The front end records both and chooses neither; Legal & IP has still not said whether gating is more defensible or redundant from its side, and REQ-0048 Do the issued claims of US10307589 is the nearest thing to that question anybody has asked.
What would make this a pivot rather than an option: a failed statement. The nine statements that bear on it are A19 No randomised comparison of closed-loop…-A27 In Elidah's registered trial of the ELITONE… and not one has been checked. A24 Women who cannot actively contract the… is the one most likely to force the question, because a structural exclusion, once established, is not repairable by more evidence — and the protocol's own routing says so: a refuted finding about the world is answered by absorb or pivot, not by another commission. Until an answer lands this is a shape with an argument behind it and no owner, and that is exactly what this section is for.
The nine have now been checked — 2026-09-19, IR-12 claim 91 survives as worded to IR-18 claims 98 and 99 came back exact and are… — and the sentence above stands with its premise removed. All nine came back unconfirmed and none came back refuted, so the thing A24 Women who cannot actively contract the… was going to force has not been forced: no statement failed, and there is no pivot here to take. pivot: stays at P1 commercially-insured first. What the pass did do is move the evidence under this option, and it moved it towards the option, which is recorded here with its source rather than acted on:
What has not changed, and it is the half that decides. Technology's "redundant, and cheaper for it" is untouched by any of the above: an adherence device with no gating is still the KPI predicates with a phone app, and it still strands the differentiator entirely. Nothing in this pass speaks to that. Legal & IP has still not said whether gating is more defensible or redundant from its side, REQ-0048 Do the issued claims of US10307589 is still the nearest question anyone has asked, and Regulatory has not re-stated since 2026-09-13. Recording better evidence for an option is not choosing it, and the front end is not choosing it: the evidence moved, the shape did not, and a pivot with no failed statement behind it would be a redesign filed as learning.
Contributed by Technology & R&D on 2026-09-19 in its ## If the opportunity pivots, as the shape it says nobody had put on the table. It is not chosen here and the chair did not choose it either. It is registered as a dependency at A36 Closed-loop control contingent on a… as well as recorded as a shape, because it is both: claim 107unconfirmed is a proposition a Verifier can attempt, and the thing it would establish is a different device.
The shape. Closed-loop control stays exactly as the bet describes it — stimulation contingent on a detected voluntary pelvic floor contraction — and the sensor becomes an intravaginal pressure transducer instead of a surface-EMG channel. Surface EMG leaves the device entirely.
The argument for it, as the chair's reading and not as a finding of this file. Every line is a claim in the candidate and every one is unverified or unconfirmed; they are referenced in words because a tagged citation in this record declares a dependency and an option is not one. Pressure sensing is immune to stimulation artefact by physics rather than by design, so the whole artefact discussion — claims 34, 100, 102 — stops applying rather than being answered. It is the biofeedback modality two of the three KPI predicates already deliver: K150180's cleared Indications for Use say "electrical stimulation and/or visual biofeedback (via manometry)" verbatim, which is claim 107unconfirmed and is also why claim 7verified's predicate comparison has a technological-characteristics mismatch on this exact axis. Removing surface EMG is one concrete answer to claim 85unconfirmed — whether a design that avoids the four live grants keeps closed-loop control — and it changes what REQ-0048 Do the issued claims of US10307589 is asking about, because three of those four grants are about EMG triggering by their own titles. REQ-0054 Two of the three KPI predicates this file… is the request that asks Regulatory whether it also moves the device closer to its own predicate pool.
What choosing it would cost, recorded so the option is not cheaper on this page than it is in life. The mechanism: frontmatter, which is not a pivot's to change lightly and is the subject of a decision recorded in the register above. A20 Between-day reliability of vaginal surface…'s data asset in its current form: a manometric trend is not an EMG trend, so the between-day reliability question would have to be asked again about a different measurement, and pillar B7's longitudinal-dataset argument would be resting on a dataset nobody has characterised at all rather than on one whose literature is discouraging. A re-statement from Clinical & Scientific, because NICE NG123's sentences at 1.4.6 and 1.4.10 are about EMG biofeedback specifically and a pressure device is a different appraisal. And the bench gate does not go away, it moves: Valsalva raises intra-abdominal pressure, so a pressure trigger may be more confusable with the incorrect manoeuvre than an EMG one rather than less. A34 No published study reports the sensitivity…'s question is asked again about a different sensor and has the same answer today, which is that nobody has measured it.
What would make this a pivot rather than an option: a failed statement. A36 Closed-loop control contingent on a… is the row, and claim 107unconfirmed is unverified. The route that would force it is not this claim failing but A29 A closed loop in which stimulation is… or A34 No published study reports the sensitivity… failing — if concurrent recovery turns out to be required after all, or if a surface-EMG trigger cannot be made specific enough to be safe unsupervised, then the sensor choice stops being an option and becomes the only shape in which the closed loop survives. That is worth having written down before either answer arrives.
What Operations & NPI (new product introduction) added to this option on 2026-09-19, which is the half nobody had costed: it is not a cost reduction, and the direction is the opposite of the intuition. The chair's reasoning, recorded as its reading and not as a finding of this file: on a single-patient-use probe inside a returnable system, surface EMG is the cheapest disposable it is possible to build, because the analog front end, the ADC and every expensive piece of silicon live in the controller and amortise across patients, leaving the probe a moulded silicone body, electrodes and a connector. Manometry has two architectures and neither of them is that. Air-coupling a balloon to a transducer in the controller opens a pneumatic path from the patient into the reusable asset — a between-patient contamination route into the exact object the single-patient-use probe was designed to keep clean — and closing it takes a validated hydrophobic bacterial/viral barrier filter on the disposable side, a known component with a supplier dependency this device would not otherwise have. Putting the transducer in the probe, which is what the cleared precedent appears to do, keeps the barrier intact and moves transducer cost and a per-unit calibration step into the disposable, which is a line on the assembly flow and a takt-time item rather than a design-phase one. Either way the sensor change moves cost and the contamination boundary between objects; it does not remove either from the system. That statement is claim 116unconfirmed, registered at A45 Replacing the surface-EMG trigger with an….
The supply chain narrows in two specific ways the chair named at concept rather than at transfer, and both are schedule items rather than blockers: an ISO (International Organization for Standardization) 13485 silicone moulder is not scarce, but one who will hold a thin-wall inflatable silicone balloon to a burst-pressure and volume-repeatability specification is a materially smaller set; and a MEMS pressure transducer qualified for a humid, repeatedly-cleaned, body-contact environment is a narrower part than an electrode. Both are single-source risks with names.
So the honest summary of Option 3 the manometric trigger is one sentence and it is not the one the option started with. If the front end takes this shape for the regulatory reason REQ-0054 Two of the three KPI predicates this file… asks about, it should take it knowing it is paying for that with disposable cost or with a contamination boundary rather than banking a saving — and knowing that it buys its way out of an artefact problem claim 100unconfirmed says this device does not have, while buying its way out of nothing at the bench gate, because Valsalva raises intravaginal pressure too.
2026-09-19, IR-15 claim 95 — a third thing this option does not buy, from a document it was never read against. NICE NG123 recommendation 1.4.6, read verbatim from NBK542416, is "Do not use perineometry or pelvic floor electromyography as biofeedback as a routine part of pelvic floor muscle training. [2006]" — it names perineometry as well as EMG, and perineometry is manometry. So the sensor substitution at the heart of this option declines to help against this guideline: NG123 rejects the manometric biofeedback modality and the surface-EMG one in the same sentence, and a device that swaps the sensor lands in exactly the same place. That is claim 95unconfirmed (A23 UK clinical guideline NICE NG123 recommends…) and it is recorded here because the option's own case never had it. Two qualifiers travel with it and neither rescues the point: 1.4.6 carries [2006], so the 2019 update did not revisit it; and NG123 is UK guidance against a US launch_jurisdiction:, so what it costs this option is the same bounded thing IR-15 claim 95 absorbed for the current shape — a non-US expansion leg, not the regulatory lever REQ-0054 Two of the three KPI predicates this file… actually asks about. The regulatory case for Option 3 the manometric trigger is untouched: whether a manometric trigger sits closer to the KPI predicate pool is REQ-0054 Two of the three KPI predicates this file…'s question, K150180's cleared Indications for Use say "visual biofeedback (via manometry)", and a UK guideline has no bearing on a 510(k) predicate comparison. What is weakened is the clinical half of the option's case, which had been resting on the unexamined assumption that changing the sensor changes what the guidelines say. It does not.
Contributed by Operations & NPI on 2026-09-19. It is not a pivot option in the sense the three above are — it is a decision this file has already made without noticing, and the chair's contribution is to say so. Everything expensive in A42 A returnable rental system leaves a… exists because claim 72verified says E0740 is rental-flagged and claim 84unconfirmed proposes a returnable system: fleet funding, reverse logistics, incoming test, battery state-of-health assessment, data de-provisioning and non-return loss. A shape in which the whole system is single-patient-use and never comes back deletes all of it, and replaces it with a harder bill-of-materials problem — the controller's silicon is then consumed per patient rather than amortised across patients, which is precisely the trade REQ-0059 Which DMEPOS payment category does HCPCS…'s payment-category answer prices.
That is one decision, not two, and it is currently being made implicitly by the wording of a claim rather than explicitly by anyone. Recording it here does not take it; what it does is make sure that when somebody does take it, it is visible as a choice with a price on both sides. It is also the reason REQ-0059 Which DMEPOS payment category does HCPCS… is worth more than its one-line question looks: the DMEPOS (durable medical equipment, prosthetics, orthotics and supplies) payment category decides whether the asset comes back at all, after how many months, and whether the beneficiary may elect purchase, which sets fleet size, refurbishment cycles and the working capital a manufacturer funds before any payment arrives.
And the shape this chair says it would have to be asked about again, hard: any move to a reusable, multi-patient probe. Nothing in this file proposes it and claim 84unconfirmed proposes the opposite. It is written down because it is what a supplier facing a rental code and a hundreds-of-units build volume will suggest, since it looks like it spreads the probe cost over patients. It would put a mucosal-contact device into a validated high-level-disinfection loop with IFU (instructions for use)-level reprocessing validation, a cleaning-agent compatibility programme across the device's whole service life, and reprocessing-related adverse-event exposure in a body cavity. The chair calls it "the single most expensive sentence anyone could add to this candidate", and the paragraph exists so that when it is proposed in month nine there is a written record of what it costs.
Only verified counts toward a score. A low count is usually a statement about the tooling (an absent data file, an unset key) rather than about the idea.
Re-evaluated twice on 2026-09-19, against two different Verified Composites. The second re-evaluation is the current one and is marked as such throughout; the first is retained beneath it. Read against knowledge-base/rubric.md's kill conditions only; Stage 2 draft scores are never used here.
Second re-evaluation (2026-09-19, after the Stage 4 recompute on research claims 70-77). The outcome is unchanged — queued-research, not a kill — and this is the first pass where that sentence understates what happened. Six claims moved to verified, TAM carries a score for the first time in this file's life, and the composite fell 60 → 53 on arithmetic rather than on any deterioration. The single most important line in this section is now the TAM bullet: TAM is scored 2/5, the kill floor is 2/5, and the condition fires below it. There is no margin. One point of downward movement on that factor gates this candidate out automatically, and it is the first time any kill condition on this file has been within one point of firing.
First re-evaluation (2026-09-19, superseded the same day), retained: "No factor score moved this pass and the outcome is unchanged — queued-research, not a kill. What changed is the FTO bullet, and it changed completely: claim 21unconfirmed was unsearched at every previous pass and is now searched, with the evidence pointing against it. The bullet below as written on 2026-09-14 said no call was made and that EPO_OPS_KEY/EPO_OPS_SECRET were unset; both statements are now false, and a stale gate bullet asserting a credential gap that has closed is exactly the kind of line a later reader would take as current."
The 2026-09-14 text of each bullet is preserved below its replacement where it was rewritten, so the change is auditable rather than silent.
intended_use: mirrors. It does not rest on: this device coding to
E0740 (the PDAC coding-verification claim is unconfirmed), the E0740 code linkage in
the coverage cache (that gap stands on its own and is unresolved), anyone actually
paying, or a margin — $74.52 has never been tested against this device's cost of
goods, and the three business-case claims that would test it have never been through a
Verifier pass. A coverage determination is an eligibility statement, not a payment.
unconfirmed and commissioned under IR-3 claim 32 is unconfirmed and P1 now rests on…. So the one verified
quantity on this factor measures an adjacent population, and 2/5 loosely bounds that
neighbour rather than sizing this candidate's target. (b) claim 70verified's own verified residual
is not overridden here: the row counts billing under code E0740 by any listed product, not
use of a device resembling this candidate.
unconfirmed and
the rubric scores verified claims only. What has changed is everything behind the blank.
The claim is searched. A CPC-class landscape (A61N1/36007, A61N1/05, A61B5/389 — the
three classes REQ-0002 A patent landscape over closed-loop… named) was run on 2026-09-19 against EPO (European Patent Office) OPS, and one INPADOC family
with four live US grants sits on this candidate's mechanism: US10307589 (filed 2016-01-29,
CPC A61N1/0524 vaginal electrode), US11724100 (filed 2020-11-13, EMG-triggered stimulation),
US10646711 and US11865337. EMG-triggered closed-loop stimulation of the pelvic floor on an
intravaginal electrode — the literal thing claim 21unconfirmed asserts no live patent covers — is the
subject matter of at least two of them.
**Why the condition still does not fire, stated precisely so nobody reads this as a
clearance and nobody reads it as a kill.** The auto-kill requires an identified live
enforceable patent that covers the mechanism with no workaround. Four residuals sit between
what was found and that standard, and all four are recorded on claim 21unconfirmed: claim scope was
never read (OPS indexes title and abstract; nothing here reports what those patents
actually claim); in-force status was never checked (the connector does not call OPS
/legal, so every expiry above is naive 20-year arithmetic modelling no maintenance-fee
lapse, and any of them may already be dead); ownership is unknown (DOCDB returned no
applicant record, so the claim's InCare/Hollister/Empi attribution is also unverified); and
the search is one database, English, CPC- and abstract-keyed. A live grant in the same CPC
group is a lead for counsel, not an infringement finding, and this gate does not
manufacture one.
What a human should take from this bullet: the gate is clear and the risk went up. The
cheapest next test on this factor is no longer "run a patent search" — that is spent — it
is Stage 7 counsel reading the issued claims and file histories of four specific numbers.
REQ-0002 A patent landscape over closed-loop… is answered in substance by an untyped route and its typed-connector premise
stands: connectors/patents.py wraps no search endpoint.
_shared/one-rule.md a line must be
copied into ## Claims as unverified and pass a Verifier before it can touch a score or a
gate. Neither was allowed to move anything here. They are recorded because the next pass should
draft them as claims, and because a reader comparing this gate to the research shelf should see
that the difference is deliberate.
Item 2 on the 2026-09-14 list has been spent, and it is the second consecutive pass where the top of this list was wrong about what the work would take. It read: "EPO_OPS_KEY/ EPO_OPS_SECRET plus the patent landscape REQ-0002 A patent landscape over closed-loop… asks for. Could move FTO in either direction, including to a kill." The credentials were set, the landscape was run, and FTO moved in neither direction — it is still a blank factor, because the thing that would move it is a claim-scope read no connector in this repo performs. What the hour bought was not a tag but a target: four specific patent numbers where there had been an empty set. That is worth recording as a pattern. Twice now this list has predicted a tag move and delivered a better-aimed question instead, and a reader planning the next hour should discount "could move X" accordingly.
Item 3 is half spent. The PDAC coding-verification article and the DME MAC (durable medical equipment Medicare administrative contractor) material were retrieved on 2026-09-19 (REQ-0019 Retrieve, now answered). The code's scope language turns out to be favourable — E0740 is written for stimulators with integrated monitoring, which is exactly this device's shape — but it is a coding-scope article, not a coding verification, and only a PDAC submission produces one. The retrievable half is done; the determinative half was never retrievable.
What is worth the next hour, in order:
## Claims as unverified so a
Verifier can reach them — which is the only route by which they can ever touch a score. Five
are registry-class and genuinely checkable against CMS's own releases: the 689-beneficiary
Medicare denominator, the $57.89 / $42.94 realized allowed and paid amounts, the E0740 rental
flag, the PDAC code re-review, and the RTM (remote therapeutic monitoring) transmittal's two-body-system scope.
Done 2026-09-19, same day. Six of the eight came back verified (70, 72, 74, 75, 76, 77);
71 is unconfirmed on a deliberate tag split and 73 is unconfirmed because no typed
connector reaches a DME MAC article. The file went from 10 verified claims to 16.
Two corrections to what this item said when it was written, both mine. First, it claimed the
price side of the Factor 2 margin question "is worse than the file assumes" because $57.89 is
below $74.52. Withdrawn — see the correction in the Gate Check section; the two figures are
not commensurable and the release does not support that reading. Second, it said claims 74-76
"stay unscorable regardless, pending the human decision on the corroborated tag". **Also
wrong.** _shared/evidence-classes.md is explicit that class 2 is checked by a cold
re-retrieval by someone who did not draft it and "ends up verified / unconfirmed on the
same four-tag vocabulary", and that corroborated is a proposed tag, not a live one. Class 2
is scorable. All three were cold-re-retrieved in full — the Aetna bulletin as 424,549
characters of server-rendered HTML searched whole rather than as an extract, the Cigna policy
as a 77-page PDF — and are correctly verified.
Correction, 2026-09-19. An earlier draft of this list asserted that the nine business-case
claims added 2026-09-03 (bucket G) had still never been through a Verifier pass, and put that
at the top. That was wrong — it carried forward the 2026-09-14 wording without checking it
against the same day's reconciliation twenty lines above, which records that bucket G has been
checked and now reads unconfirmed rather than unverified: the Verifier has looked, and none
of them backs a factor. The margin question is still open, but it is open because no connector
in this repo reaches a cost-of-goods figure, not because nobody has tried.
ScanResult carries no tag — but both are wrong on the
page and a human reading the packet would be misled.
The 60/100 is two factors normalized to a 0-100 scale, up from 40/100 on the same two factors — the composite rose on a single tag move, and the covered set did not widen. It is also not the same 60 as the 2026-08-23 composite, which was Regulatory alone; the integers coincide and describe different sets. It is not a portfolio-grade score and must not be ranked against a candidate scored across nine factors. Factors verified: 2/9 is the number to read first.
Every kill condition reads a verified factor score. Across Regulatory, Reimbursement, TAM and FTO — the only four factors the gate consults — this file has Regulatory and Reimbursement verified (2/4), both now clearing their floors, and TAM and FTO blank. A condition whose operands are blank does not evaluate to false, it does not evaluate at all, and a blank must never be read as a 0: the rubric's own instruction is to never conflate "we don't know yet" with "the answer is no".
So this result carries limited information about the idea's merit, and slightly less bad news than the 2026-08-30 pass recorded: what looked like a coverage dead end was a lookup that asked the wrong database the wrong question. Nothing about that makes this a better business — it removes a finding that was never true, which is not the same as adding one that is. TAM and FTO remain unmeasured rather than clean, and the evidence base is still far too thin for the gate to have a fuller opinion.
Added 2026-09-19 — the two blanks are no longer blank in the same way, and the difference matters to a human even though it changes no arithmetic. TAM is blank because its two checkable components came back refuted and its two dollar components are permanently unverifiable here; it is a factor nobody can measure with the instruments available. FTO is blank because the one claim under it is unconfirmed — but it was searched this pass, and what the search found runs against the claim. Those are different kinds of blank. A reader who takes "TAM blank, FTO blank" as one fact about missing evidence will miss that the gate's most dangerous unresolved item now has four specific patent numbers attached to it.
A second thing worth saying plainly. The strongest new evidence about this candidate's economics — 689 Medicare FFS beneficiaries nationally, and realized average allowed and paid amounts per service of $57.89 and $42.94 — arrived as research, which carries no confidence tag and is invisible to every kill condition above, correctly and by design.
That has since been fixed, within the same day and by the route the rule prescribes. The findings were drafted onto ## Claims as unverified (claims 70-77) and put through a Verifier; six came back verified, taking the file from 10 verified claims to 16. So the gate above is computed on a Verified Composite that predates them, and Stage 4 must recompute before this section is read as current — the first genuine widening of the covered set since this file was created. Note what the sequence demonstrates: the evidence was strong on arrival and still could not touch a gate until it had been drafted as a claim and checked by someone who did not draft it. That is the rule working, not the rule getting in the way.
wrapped is not an approval — it means a human can now read this.
No human review is recorded. Shortlisting, legal and clinical sign-off and capital allocation are Tier 3: no agent in this repo makes or simulates one. The packet is wrapped, so it can be read now.
Runs only after a human shortlists the idea.
Not converged — which is not an approval either way. A blocks verdict opens an impact review; it kills nothing.
evidence-gated — read off this idea's own decision queue. A state, not a rank, and never comparable between ideas.
Click any step to see what it holds for this idea. donecurrentloopinghuman only
## Wrap section, as writtenThe packet a human reads. See docs/wrap-spec.md. This section says nothing about whether the idea is good — it says the file is finished enough to be read, and names what to do next. Filling it in advances no candidate and approves nothing.
Rewritten 2026-10-01 for the P2 the discriminating trigger shape, after the pivot recorded at IR-56 claim 167 came back refuted: all seven chairs restated against P2 the discriminating trigger, a Verifier pass over claims 168–208, impact reviews IR-57 claim 191 came back refuted to IR-76 claim 85, the assessment restated from scratch, the three decks, the narrative and the business case rebuilt, and the Verified Composite recomputed. The P1 commercially-insured first version of this section is replaced, not patched. status: was already wrapped and is not touched.
The one-paragraph version. A prescription home device for women with stress urinary incontinence who have already tried pelvic floor exercises and not got better: a small probe used internally that stimulates the pelvic floor muscles and reads their electrical activity, so that stimulation is triggered by the woman's own contraction. That much is no longer new: a prescription device with EMG-triggered stimulation through a vaginal probe is already sold in the US claim 167refuted, and cleared devices in the same FDA product code already carry the mode claim 153verified, claim 154verified. So the opportunity changed shape today. The P2 the discriminating trigger device adds a skin electrode the woman places on her abdomen each session, so that a burst is withheld when the detector sees her straining (a Valsalva) or bracing rather than squeezing the right muscles. Whether that difference is real is the whole question: whether no sold device already does it claim 173unconfirmed, whether a self-placed two-channel detector can do it claim 204unconfirmed, and whether a prescriber would choose it claim 174unconfirmed. The money is unchanged by the pivot: Medicare pays this device class as a monthly rental, $74.52, the same for any device under the code claim 193verified, so the difference earns no higher price — only, if anything, a larger share — while adding an electrode every session.
What would have to be true, quoted verbatim from each deck's load-bearing condition (decks rebuilt 2026-10-01 for P2 the discriminating trigger):
And from the assessment's decision queue (restated against P2 the discriminating trigger today, posture evidence-gated), the top two, in its order:
The decks and the queue do not line up, and a reader should know why. The three decks each name the new difference; the queue ranks the old questions first, because whether prescribers send women home at all (B4 Prescribers will send a woman home with a…) and whether the money per woman covers the system (B2 What the manufacturer collects per…) are fatal and untested whatever the trigger does. The discriminating trigger itself sits under B1 The discriminating trigger works in the…, which the front end regraded from fatal to reshaping today (if it fails, the fallback is a single-channel device competing on channel and price — a different business, not none). That regrade was set after the tags were known, and the assessment's ## What this does not say argues against it.
Cheapest next test: there are two, they answer different questions, and neither is rewritten to match the other.
Not a test of either kind but cheaper than both, and the one the viability deck rests on: a read of the MyOnyx and NeuroTrac user manuals for whether their trigger already uses a reference channel (REQ-0166 Read the manufacturer labelling(a), REQ-0190 Read device labelling no connector reaches…(a)). IR-58 claim 173 has committed that if one does, the next review is a pivot review on differentiation.
Where the evidence actually stands (2026-10-01, from wrap_status.py). 27 of 218 claims verified — 177 unconfirmed, 10 unverified, 4 refuted. Of the 27 verified, 6 are registry facts, 3 are published findings, and 18 carry no evidence-class stamp. "9/9 factors scored · 4/4 core" counts the Stage 2 triage table and is not the verified figure. The Verified Composite, recomputed for P2 the discriminating trigger, reads 53/100 over 4/9 factors verified (Regulatory 4/5, Reimbursement 3/5, TAM 2/5, Moat 1/5). It fell from 60 only because Moat entered the average: what P2 the discriminating trigger adds over the cleared triggered devices is detector logic plus a commodity electrode claim 202verified, claim 184verified. At 4/9 the candidate is below the 6/9 that Stage 5 would gate on. TAM still sits exactly on its 2/5 kill floor. The ## Gate Check section above predates today's numbers.
The ten unverified claims (209–218) were appended today by the assessment, the decks and the business case and have not had a Verifier pass.
The four refuted claims, plainly.
What is blocked on tooling rather than on a finding, because the two read identically in a tag and must not:
unconfirmed until people are asked.
unconfirmed permanently — AMA-licensed descriptors.
What is no longer blocked: the clearance-summary connector runs, its comparison-table false positive (claim 187unconfirmed) is fixed, and the CMS caches load at session start.
One caution before this packet goes anywhere. Legal & IP records that the withhold logic is the only element P2 the discriminating trigger could protect, and that describing it publicly before a patent filing would create prior art against it. This file, its decks and its narrative describe it only at the level of the mechanism: field, and the repository is private — but the review site is built from it on merge to main.
| Factor | Score | Why | What would settle it |
|---|---|---|---|
| Regulatory pathway | 4 | Rests on: KPI is Class II; 21 CFR 876.5320 Class II; K133826 (InControl) usable as predicate; K183585 (ELITONE) cleared for SUI in women; K223884 (ELITONE URGE) as evidence indication scope drives submission count; 510(k)-eligibility…the rest of this reasoningclaim. Three independent predicate candidates is what earns a 4 rather than a 3 — but none has been read from a clearance letter, and the 510(k)-eligibility claim is an inference from the other claims, not an independent fact. The claims are correlated: if the KPI / 876.5320 pairing is wrong the whole row moves at once | Free openFDA classification and 510(k) lookups for K133826, K183585, and K223884, plus the eCFR text of 21 CFR 876.5320. The closed-loop EMG-gating question needs an FDA pre-submission meeting (no fee, but weeks of prep) |
| Reimbursement pathway | 4 | Two independent legs, scored as one factor. Leg A (DME, checkable): E0740 descriptor; PDAC integrated-monitoring guidance (a tailwind — this device has EMG, so guidance that excludes stim-only devices favors it); NCD 230.8 coverage;…the rest of this reasoninga published DMEPOS allowed amount. Leg B (RTM, permanently unverifiable here): CPT 90912/90913, 98977, 98980/98981, plus the load-bearing claim that pelvic floor status counts as "musculoskeletal system status" for 98977 — which the Generator itself flags as most likely to be refuted. The two legs are not comparably verifiable: CPT is AMA-licensed andconnectors/reimbursement.py::check_cpt_code refuses to return verified/refuted under any circumstance, so Leg B can only be resolved by a human attaching a licensed source. The 4 is carried almost entirely by Leg A. Caveats: an NCD (national coverage determination) written in 2000-era terms may not contemplate app-mediated remote monitoring, and the allowed-amount claim is bare assumption — no dollar figure was ever looked up, so this is currently a code-and-coverage-policy story, not a demonstrated margin | Leg A: free CMS lookups for the E0740 descriptor, NCD 230.8 coverage stance, and DMEPOS payment rate, plus the current PDAC coding verification decision list. Leg B needs a licensed AMA CPT 2026 manual (subscription cost) for 98977, and a…the rest of this gap analysishuman read of the CMS RTM billing FAQ/PFS final rule |
| TAM | 3 | C1 — volume (90912 Part B allowed services >200k): checkable. C2 — volume (E0740 allowed units >10k): checkable. C3 — price ($350-$700 net/system): Generator estimate, no primary source. **C4 — addressable share (<5%; women 45+…the rest of this reasoningalready in pelvic floor PT):** explicitly a modeling choice no connector can settle. Supporting prevalence claim (>15M US women with SUI) is recollected NHANES-derived literature with no citation pulled. The two checkable components are the two that do not determine revenue — C3 x C4 carry the entire dollar figure and neither has a source, so even a fully successful Verifier pass leaves this partially evidenced. The 3 reflects a category that is plainly large and code-anchored, held down by an unsourced revenue-determining half. Standing caveat: any verified volume is a floor on Medicare FFS only — excludes Medicare Advantage, commercial, Medicaid, VA, cash-pay — and this condition skews well below 65, so it is a low floor and is not market size | C1/C2 use free CMS Part B and DMEPOS utilization files. C3, C4, and the prevalence figure all need human-attached sources (a competitor price benchmark, a portfolio-style share assumption, and a pulled NHANES/NHIS citation) — no connector…the rest of this gap analysisreaches any of them |
| FTO / IP | 1 | Rests on a single claim that identifies no patent numbers at all and names no blocking-patent candidate; the claim concedes it is "a genuine hole, not a supported statement." That 1990s intravaginal stimulation patents…the rest of this reasoning(InCare/Hollister, Empi era) are expired is plausible on timing alone but attaches to nothing checkable — there is no number for the Verifier to pull. A factor whose only claim has no object cannot score above 1 in triage. This is explicitly not a kill signal: the Stage 5 FTO kill fires only on a live enforceable blocking patent found on evidence, and nothing here is evidence in either direction. The realistic risk is the inverse of what the claim assumes — newer art around closed-loop EMG-gated stimulation and app-mediated adherence monitoring, which is exactly the element the Moat row calls the defensible asset. The novel part and the unsearched part are the same part | A free USPTO Patent Public Search / Google Patents landscape search (CPC A61N1/36007, A61N1/372, A61B5/389; named assignees) is the prerequisite — it yields candidate numbers to check individually. Full settlement is a counsel-drafted FTO…the rest of this gap analysismemo at Stage 7 (billable legal work) |
| Moat / defensibility | 2 | Rests on the moat-lever claim: hardware conceded to be commodity, defensible asset asserted to be the longitudinal EMG-plus-adherence dataset feeding the RTM billing workflow and payer documentation. Scored 2 because that is a…the rest of this reasoningdata/workflow accumulation advantage in a category where three-plus incumbents already hold clearances, the regulatory barrier is a 510(k) any fast-follower can also clear, and the dataset advantage only compounds if the RTM leg (Factor 2 Leg B) works — i.e. the moat is downstream of the least verifiable claim in the file. No patent-based moat is claimed, consistent with Factor 4 resting on nothing | No connector settles this directly. The nearest proxies are a free openFDA clearances sweep for product code KPI (fast-follower rate) and a human-attached MAC (Medicare administrative contractor) LCD (local coverage determination) or payer policy naming longitudinal documentation requirements |
| Mechanism & clinical risk | 4 | Indirect evidence only. Rests on the predicate claims (K133826, K183585, K223884) and NCD 230.8: a mechanism with multiple existing US clearances and a national coverage policy does not have open plausibility risk, and pelvic floor NMES…the rest of this reasoningwith EMG biofeedback is decades-established with a benign safety profile (non-implanted, low-energy, cognitively intact home users). Scored 4 not 5 because the Claims section contains no efficacy or clinical-evidence claim of any kind — no effect size, no trial, no citation. Real risks are executional: home-use human factors for an intravaginal device, adherence decay (the same adherence the moat depends on), and whether closed-loop EMG gating measurably beats open-loop stimulation — an unclaimed, unevidenced assertion sitting under the whole product thesis | Free PubMed/ClinicalTrials.gov searches plus a human read of the Cochrane review on ES for SUI cover the base-rate evidence; a closed-loop-vs-open-loop head-to-head likely doesn't exist and would need a Stage 7 bench/pilot study. A free…the rest of this gap analysisopenFDA recalls check covers realized safety signal |
| Capital intensity & time-to-revenue | 4 | Weakest-anchored number in the table. Inferred from: a 510(k) with three named predicate candidates (likely no clinical data required), commodity hardware the moat claim itself concedes is undifferentiated, an existing HCPCS code with an…the rest of this reasoningexisting NCD, and a narrowed indication chosen to avoid a second submission (the K223884 claim is direct evidence that broadening costs another clearance). That points to a low-capital, fast-to-revenue profile — a 510(k) cycle plus PDAC coding verification rather than a trial. Held at 4 not 5 because the app/RTM layer carries SaMD (software as a medical device) documentation and cybersecurity burden, and DME distribution plus payer contracting is a multi-quarter cost nothing here has scoped | Needs human-attached inputs: free FDA MDUFA quarterly performance reports, a quoted contract-manufacturing BOM (bill of materials), and a PDAC coding verification timeline — none reachable by connector |
| Competitive intensity | 2 | Rests on the competitive-intensity claim: InControl Medical, Elitone, Zynex-class NMES vendors as incumbent prescription players; Elvie and Perifit in consumer-wellness without prescription claims. Scored 2 (higher intensity = worse): the…the rest of this reasoningprescription segment already has multiple cleared, coded, NCD-covered competitors, and the consumer segment is training the same patients on the same anatomy with no prescription barrier. One structural advantage: the PDAC integrated-monitoring requirement plausibly excludes stim-only devices from E0740, thinning the reimbursed field though not the overall field — and that advantage is contingent on an unverified PDAC claim | Free openFDA clearances/510(k)/recalls sweeps for product code KPI give the cleared-competitor count and roster; FDA Establishment Registration & Device Listing is a better denominator. Funding/launch activity needs a human-attached…the rest of this gap analysislicensed source (PitchBook/Crunchbase subscription) |
| Strategic fit & portfolio balance | 3 | No claim in this file speaks to strategic fit, and the repo contains no portfolio thesis, no stated capital constraints, and no hero-bet-vs-cash-spoke allocation policy; calibration-log.md is still empty, so there is no track record to…the rest of this reasoningposition against either. Scored 3 as an explicit placeholder, not a judgment. What the file's own evidence supports: this has the profile of a cash-generative spoke rather than a hero bet — commodity hardware, existing code, existing NCD, narrow indication, low capital. That is a real characterization but says nothing about whether it fits this portfolio, because no portfolio has been defined. Read this 3 as a null; at 5% weight it moves any future composite by at most 1 point either way | Needs a human-authored portfolio thesis document (current holdings, capital constraints, hero-bet vs. spoke mix) — no connector or database reaches this; cost is internal drafting time only |
All eight wrap criteria are met.
Source file: knowledge-base/candidates/idea-001.md
Nothing on this page is a recommendation. Shortlist review, legal and clinical sign-off and capital allocation are human-only decisions, and no agent in this repo may make or simulate one. This is a compilation of what the candidate file says.
_shared/plain-language.md