hammer-building-the-evidence Scope: world hammer-coverage-intelligence vintage 2026-10-01 Revision SHA-256: 4ca92af634863c598fa9829668e2a208a1289607bc5c47e2d97aa4e6a0a71344 --- name: hammer-building-the-evidence description: Use when somebody has a commercial goal and needs evidence that survives the buyer's own scrutiny: pricing a service line, pitching a payer, sizing a market, arming a contract negotiation, showing an association's members what their data is worth. Carries the order of operations from a goal to a defensible claim, how to find what data exists before spending a call, the industry's own statistical toolkit mapped to the estimator that serves each, the confounds that have killed real findings in this corpus with what each one cost, how to read a refusal as the next step rather than a wall, and a worked example from a live commercial question to the evidence that answered it. This is the method; the persona skills route the question and hammer-writing-the-deliverable assembles the document. --- # hammer-building-the-evidence Somebody has a goal. Win the contract, get the code covered, price the service line, convince a plan to buy. **They do not need to be told what this world cannot do. They need a path to evidence a buyer will accept**, and the fastest way to lose is to build a beautiful analysis on a figure the buyer's own analyst can dismantle in one question. That is what this world is for, and it is why the checks exist: not to stop the work, but to stop the work from being demolished in the room. ## The order of operations **1. Write the buyer's test, before touching data.** Not your claim: **the sentence the buyer has to accept.** A plan's actuary will accept "this reduces admissions per thousand" and will not accept "our device is better". A hospital's contracting lead will accept "your case rate is below the market's middle for this DRG" and will not accept "you are underpaid". Writing the test first tells you what evidence class you need, and evidence class decides everything after. **2. Find out what exists, before spending a call.** `knowledge_navigate` reads the source register, and its three states are the whole point: - **`shipped`** means the bytes are here and a named tool reads them. This is what you can measure. - **`available`** means the source exists upstream at a named vintage and this world does not carry it. **This is a fetch, not a wall.** It is a week of work, not a dead end, and knowing the difference is worth more than most answers. - **`absent`** means it was never fetched and the reason is recorded, usually a licence or a data use agreement. That is a commercial conversation, not an engineering one. A question that has no compartment and a question nobody packed the data for look identical from the outside. The register is what tells them apart. **But the register can understate its own world, so read it against the tool surface rather than instead of it.** Measured 2026-08-29: the hospital standard-charge corpus is registered `available` with an empty `carried_in`, and its note reads that no tool here reads it, while `filed_charge` reads it and serves tens of millions of rows from it. An analyst who trusted the register alone would have concluded this world cannot price an inpatient admission and stopped. **So the sequence is: read the register for what is absent and why, and read `hello_world` and the tool descriptions for what is actually served.** Where the two disagree, the tool surface is what answers a call, and the disagreement is a defect worth reporting. **3. Pick the grain the buyer argues at.** This is where most analyses die. A payer argues at member-months and PMPM. A hospital argues at the case and the DRG. A device company wants to argue at the procedure and is usually forced to the episode. **Get the grain wrong and every figure downstream is unanswerable**, however well computed. **4. Pull the figures, and read what rides beside them.** Every cell in this world carries `n`, a basis, and its thinness figures. Those are not decoration: they are the answers to the questions the buyer's analyst will ask, arriving before the question. **5. Check the confounds, on the list below, before you believe your own finding.** **6. Say what you can say.** Then, and only then, the document. ## The statistical toolkit, and which tool serves it Industry has a small set of moves it accepts. Each has an estimator here, and each estimator exists so the arithmetic happens behind a receipt rather than in a chat turn. | the move | when the buyer expects it | the tool | |---|---|---| | what a payer publishes it pays | any pricing or reimbursement claim | `rate_position` | | the same across codes under one receipt | comparing codes so the comparison is checkable | `rate_comparison` | | how a code splits professional against technical | anything about who bills what | `fee_components` | | which code pairs may not be billed together | building a billing model | `stack_check` | | who may furnish and interpret a code | market access, staffing models | `interpretation_requirement` | | what a hospital filed for an admission | contracting, service-line pricing | `filed_charge` | | headcounts and modelled prevalence | sizing any population | `cohort_figures` | | signal against noise on a small denominator | any per-provider or small-cell claim | `measure_reliability` | | how much of a change was the mean reverting | any before-and-after intervention claim | `regression_to_mean_expectation` | | positive predictive value at a real prevalence | any screening or find-the-patient claim | `screening_yield` | | a deferred cost against an avoided one | any savings claim, ever | `deferred_versus_avoided` | | what a quality withhold earns back | value-based contract modelling | `withhold_earnback` | | where a provider sits in a ranked programme | HRRP, Stars, any tournament | `tournament_position` | | a competing event that eats the benefit | mortality, dropout, any long horizon | `competing_event_coupling` | | published association, one record at a time | clinical framing | `prognostic_association` | | randomised effect, kept separate on purpose | efficacy framing | `treatment_effect` | | a customer's own numbers, declared | any model built on their book | `declared_inputs` | | exposure over a declared population | sizing a plan's spend | `cost_of_care_exposure` | | hard against soft savings | any submission to a plan | `savings_methodology` | **Three of these exist because industry gets them wrong constantly.** A per-provider ranking without reliability is sampling error wearing a league table. A before-and-after without regression to the mean credits the intervention with the mean reverting. A savings figure that sums a deferred cost and an avoided one counts the same dollar twice. If your case rests on any of those three, run the estimator before you write the sentence, because the buyer's analyst will run it after. ## The confounds that have actually killed findings here Each of these is measured, not hypothetical, and each one killed something that looked finished. **Billing class.** A hospital-filed against payer-filed reconciliation produced ratios that flipped hard and consistently by payer, and it read as a finding about negotiating posture. It was billing class: one payer's rows were 622 of 645 institutional, another's 562 of 662 professional, and the hospital side carried no billing-class column at all so it could not be split to match. **A cross-corpus ratio that tracks a variable the two corpora record differently is not a finding.** **Replication.** The same filing arrives repeatedly. Measured on one code: 8,200,690 rows in one tree against 4,270,330 in the deduplicated one, with the median and p75 identical in both. **An inflated `n` is not more evidence**, and every cell here says which tree it came from for that reason. **Grouper edition.** A code present at one filer and missing at another may be a code that filer's edition does not define. Absence settles nothing. **Code vocabulary.** One real code filed under two labels by different filers, 5.4x apart on the median. Splitting on the label splits one thing into two. **Small cells.** Thirty to sixty percent of small-cell variation is sampling error before anything else is said. `measure_reliability` turns that from a caveat into a figure. **Payer identity.** 240 distinct payer strings in the hospital corpus against seven slugs on the payer side. One company appears under several spellings, some entries are third-party administrators renting a network, one is a government programme. **Resolving them is an inference nobody has licensed**, and a pitch built on a bad join dies on the first question. ## Read a refusal as the next step Every refusal in this world names what would settle it. That clause is a work order. - *"no compartment holds paid amounts, adjudicated claims would settle it"* means your savings figure needs a claims partner or a pilot, and **knowing that before the submission is worth more than the figure was.** - *"an unweighted mean over this corpus's payer mix is a number about the corpus"* means declare your mix and the average becomes computable. The refusal is asking you for one input. - *"this filer did not file this code, and the filers that did are X"* means go look at X, or go ask which edition they are on. - *"available upstream at vintage V and this world does not carry it"* means a fetch, and now you can price the fetch. **A refusal that names its remedy is a shorter path than a wrong answer that has to be walked back.** Treat the list of refusals a question generates as the plan for the next round of work. ## What makes a claim survive the room Four properties, and every one of them is checkable before you send anything. 1. **It names its grain and its denominator.** "Per member per month" not "per member". 2. **It carries its `n` and where the `n` came from.** A reader who cannot tell a deduplicated count from an inflated one cannot check anything. 3. **It states its basis.** Measured, cited, or assumption, and a figure built from several is only as strong as its weakest input. Saying so first is what stops it being said to you. 4. **It survives its own confound check.** Run the list above against your finding and write the one that came closest to killing it into the document. ## A worked example The goal: a device company needs a payer to buy, and the pitch is that its test avoids downstream imaging. **The buyer's test**, written first: a plan's actuary will accept a reduction in imaging spend per thousand members at a stated confidence, on the plan's own book, with the savings method named. They will not accept "our test is more accurate". **What exists**: `knowledge_navigate` on the register. Rates for the imaging codes are `shipped`. Claims are not here at all, so the actuary's own book has to come from the customer, which is `declared_inputs`. **The grain**: PMPM against a declared population, because that is the currency the actuary argues in. **The figures**: `rate_position` on the imaging codes for what payers publish they pay, then `cost_of_care_exposure` over the customer's declared population to size the spend at risk. `screening_yield` for how many findings the test actually produces at the real prevalence, because a test's sensitivity is not its yield. **The confounds**: small cells on the rate side, and regression to the mean if any before-and-after is quoted. **What can be said**: the exposure, the published rates with their dispersion, the yield at declared prevalence, the savings method classified. **What cannot**: the savings figure itself, because Cost of Care is defined on paid amounts and nothing here holds one, so the submission names its assumption and the actuary prices it. That sentence, said first, is the difference between a submission that gets a second meeting and one that gets returned. HTTP bindings for this session: Explicit source-reference bindings: SKILL.md: https://space.hammer.ai/skills/global/building-sourced-excel-models.skill.txt?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm analysis.md: https://space.hammer.ai/skills/global/analysis.skill.txt?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm coverage-website.md: https://space.hammer.ai/skills/global/coverage-website.skill.txt?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm rate_position: https://space.hammer.ai/analysis/rate_position/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm rate_comparison: https://space.hammer.ai/analysis/rate_comparison/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm stack_check: https://space.hammer.ai/analysis/stack_check/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm screening_yield: https://space.hammer.ai/analysis/screening_yield/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm deferred_versus_avoided: https://space.hammer.ai/analysis/deferred_versus_avoided/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm interpretation_requirement: https://space.hammer.ai/analysis/interpretation_requirement/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm cohort_figures: https://space.hammer.ai/analysis/cohort_figures/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm comorbidity_edges: https://space.hammer.ai/analysis/comorbidity_edges/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm fee_components: https://space.hammer.ai/analysis/fee_components/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm filed_charge: https://space.hammer.ai/analysis/filed_charge/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm hypertension_evidence: https://space.hammer.ai/analysis/hypertension_evidence/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm prognostic_association: https://space.hammer.ai/analysis/prognostic_association/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm treatment_effect: https://space.hammer.ai/analysis/treatment_effect/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm regression_to_mean_expectation: https://space.hammer.ai/analysis/regression_to_mean_expectation/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm competing_event_coupling: https://space.hammer.ai/analysis/competing_event_coupling/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm withhold_earnback: https://space.hammer.ai/analysis/withhold_earnback/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm tournament_position: https://space.hammer.ai/analysis/tournament_position/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm measure_reliability: https://space.hammer.ai/analysis/measure_reliability/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm declared_inputs: https://space.hammer.ai/analysis/declared_inputs/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm cost_of_care_exposure: https://space.hammer.ai/analysis/cost_of_care_exposure/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm savings_methodology: https://space.hammer.ai/analysis/savings_methodology/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm hospital_variance: https://space.hammer.ai/analysis/hospital_variance/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm filer_inventory: https://space.hammer.ai/analysis/filer_inventory/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm candidate_screen: https://space.hammer.ai/analysis/candidate_screen/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm frontier: https://space.hammer.ai/analysis/frontier/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm judgement_check: https://space.hammer.ai/analysis/judgement_check/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm claim_check: https://space.hammer.ai/analysis/claim_check/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm deliver: https://space.hammer.ai/analysis/deliver/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm session_trace: https://space.hammer.ai/analysis/session_trace/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm progress_check: https://space.hammer.ai/analysis/progress_check/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm model_advisory: https://space.hammer.ai/analysis/model_advisory/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm knowledge_navigate: https://space.hammer.ai/analysis/knowledge_navigate/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm simulate: https://space.hammer.ai/analysis/simulate/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm sensitivity: https://space.hammer.ai/analysis/sensitivity/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm interval_check: https://space.hammer.ai/analysis/interval_check/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm decision_navigate: https://space.hammer.ai/analysis/decision_navigate/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm network_infer: https://space.hammer.ai/analysis/network_infer/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm hello_world: https://space.hammer.ai/analysis/hello_world/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm say_world: https://space.hammer.ai/analysis/say_world/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm Read the companion references here (these replace local file references in the source guide): Build a sourced analysis: https://space.hammer.ai/analysis.skill.txt?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm OData exploration and analytic cuts: https://space.hammer.ai/odata.skill.txt?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm Worked analysis reference: https://space.hammer.ai/skills/global/analysis.skill.txt?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm Building sourced Excel models: https://space.hammer.ai/skills/global/building-sourced-excel-models.skill.txt?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm Coverage website reference: https://space.hammer.ai/skills/global/coverage-website.skill.txt?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm Spine cohort and post-acute analysis: https://space.hammer.ai/skills/spine-analysis.skill.txt?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm hammer-building-the-evidence: https://space.hammer.ai/skills/world/hammer-building-the-evidence.skill.txt?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm hammer-cost-of-care-case: https://space.hammer.ai/skills/world/hammer-cost-of-care-case.skill.txt?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm hammer-coverage-navigate: https://space.hammer.ai/skills/world/hammer-coverage-navigate.skill.txt?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm hammer-provider-contracting: https://space.hammer.ai/skills/world/hammer-provider-contracting.skill.txt?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm hammer-provider-quality: https://space.hammer.ai/skills/world/hammer-provider-quality.skill.txt?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm hammer-savings-classification: https://space.hammer.ai/skills/world/hammer-savings-classification.skill.txt?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm hammer-vendor-evidence: https://space.hammer.ai/skills/world/hammer-vendor-evidence.skill.txt?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm hammer-writing-the-deliverable: https://space.hammer.ai/skills/world/hammer-writing-the-deliverable.skill.txt?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm Data: https://space.hammer.ai/data/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm Transcript: https://space.hammer.ai/session/?ledger=49ec6090f32b112cd7fc2d39a72e0974&snapshot=7a53af6a9beb58928d4a64dc92db09c9db65433f9909f1a10d16e302adaf86ba&workflow=spine-rtm