Skills

Appearance

Sources register ​

The evidence behind the skill's slow-rotting claims. Every claim the skill makes about the outside world that will not change week to week has an entry here; "where did you get this?" is answered from this file, in any phrasing (./the-skill → the laws).

What belongs here — and what never does. The register holds slow-rotting canon only: findings, methods and standards that age in years. Fast-rotting facts never enter — a price, a current API limit, a competitor's live feature stay fetch-at-decision-time rules, quoted with their check-date at the moment of use, never cached here to rot.

One fixed form per entry, so a wrong entry is visibly wrong: id · full citation · live URL/DOI/arXiv · archive link · licence · one-paragraph distillate (our words) · check-date · cited-by · reads-against.

Reads against is the field that stops a register becoming a pile. A register that can only answer "what do we have on X" will happily hold two entries that pull opposite ways and never say so — measured here 2026-09-10, when 2 of 20 entries carried a comparison with another entry buried inside a distillate paragraph and no field held it, so nothing could read it back. The field names the entries this one must be read beside and what the tension is, or one of two words that are not the same answer: none found — looked, and nothing here pulls against it — or not checked, which is honest ignorance. An empty field is a defect; the two words are not. It is symmetric: if A reads against B, B says so too, and --verify-reads refuses the pair that only points one way.

Cited-by names files, never line numbers. A line number is wrong the next time anything above it is edited, and every entry here once pointed at a line — in files that no longer exist. The durable form is the file that carries the claim; the id is greppable when the exact place is wanted.

Licence tiers (they decide what we may hold): free — an open licence (MIT, CC, a public standard); a copy may be carried later. copyrighted — citation + archive link + our own distillate, never the text itself. math — a formula recorded by name, which is not copyrightable. Each entry states its tier.

Upkeep. scripts/fetch-source.py builds and checks these entries: --resolve <doi|arxiv|url> prints a skeleton, --archive <url> triggers a Wayback snapshot, --verify walks every live URL. --verify runs at every release (AGENTS.md → The release ritual).

Strategy and discovery frames — the canon behind the catalogue rows ​

torres-ost · Torres, Opportunity Solution Trees ​

  • Citation: Torres, T. "Opportunity Solution Trees: Visualize Your Discovery." Product Talk; and Continuous Discovery Habits (2021).
  • Live: https://www.producttalk.org/opportunity-solution-trees/
  • Archive: not pinned yet — pin at the next audit
  • Licence: copyrighted — cite + our distillate
  • Distillate: One desired outcome → the opportunity space (customer needs, mapped before features) → solutions → assumption tests. The layer teams skip is opportunities; the failure mode is jumping insight-to-solution. The tree is the discovery spine for outcome-oriented teams.
  • Check-date: 2026-08-06
  • Reads against: not checked
  • Cited-by: process-discovery.md · catalogue.md · templates/SPEC-template.md

helmer-7powers · Helmer, 7 Powers ​

  • Citation: Helmer, H. 7 Powers: The Foundations of Business Strategy (2016).
  • Live: https://7powers.com
  • Archive: not pinned yet — pin at the next audit
  • Licence: copyrighted — cite + our distillate
  • Distillate: Seven durable sources of differential margin — scale economies · network economies · counter-positioning · switching costs · branding · cornered resource · process power. The current firm-level canon for "why does this position endure"; each power has a benefit and a barrier, and a claim without the barrier named is not a power.
  • Check-date: 2026-08-06
  • Reads against: morningstar-moats — the older taxonomy covering the same ground with less explanatory power; kept for the shared vocabulary
  • Cited-by: catalogue.md

morningstar-moats · Morningstar's economic moats (Stoffel's teaching table) ​

  • Citation: Morningstar's five-moat taxonomy (intangibles · switching costs · network effect · cost advantage · efficient scale), widely circulated via Brian Stoffel's wide/narrow/no-moat table.
  • Live: https://www.morningstar.com/investing-definitions/economic-moat
  • Archive: not pinned yet — pin at the next audit
  • Licence: copyrighted — cite + our distillate
  • Distillate: The older valuation lens: five moat types graded wide/narrow/none per company. Useful as a shared vocabulary and a checklist; weaker than 7 Powers at explaining why a moat holds. Source of the "no moat" honesty: name recognition without pricing power is not a moat.
  • Check-date: 2026-08-06
  • Reads against: helmer-7powers — same subject, and this is the older lens: a taxonomy without a mechanism. Where they disagree, 7 Powers explains why a position endures and this one does not
  • Cited-by: catalogue.md · templates/COMPETITORS-template.md

kaushik-stdc · Kaushik, See-Think-Do-Care ​

  • Citation: Kaushik, A. "See, Think, Do, Care: A New Content, Marketing and Measurement Model." Occam's Razor.
  • Live: https://www.kaushik.net/avinash/see-think-do-content-marketing-measurement-business-framework/
  • Archive: not pinned yet — pin at the next audit
  • Licence: copyrighted — cite + our distillate
  • Distillate: Intent-based content planning: audiences clustered by intent (See — largest addressable, Think — considering, Do — ready, Care — customers) with content and measurement per cluster. The modern reach where an awareness ladder feels dated; frame-level, not persona-level.
  • Check-date: 2026-08-06
  • Reads against: not checked
  • Cited-by: catalogue.md

perspective-ost-2026 · the starving-tree finding ​

  • Citation: Perspective AI. "The Opportunity Solution Tree in 2026: A Practical Guide for Continuous Discovery" (2026).
  • Live: https://getperspective.ai/blog/opportunity-solution-tree-2026-practical-guide-continuous-discovery
  • Archive: not pinned yet — pin at the next audit
  • Licence: copyrighted — cite + our distillate
  • Distillate: The commonest 2026 failure is not drawing the tree wrong but starving it: opportunity spaces refreshed quarterly at best under a weekly-moving market. A vendor's practice guide, not a study — the rung is cited, and the claim is carried as the craft's practice literature, not as measurement.
  • Check-date: 2026-08-06
  • Reads against: not checked
  • Cited-by: process-discovery.md

Persona theatre — grounding, fidelity, and the limits of synthetic audiences ​

park-self-reports · Park et al., self-report-grounded individual simulation ​

  • Citation: Park, J.S., et al. "LLM Agents Grounded in Self-Reports Enable General-Purpose Simulation of Individuals." arXiv:2411.10109 (2024; v1 was "Generative Agent Simulations of 1,000 People").
  • Live: https://arxiv.org/abs/2411.10109
  • Archive: http://web.archive.org/web/20260726212521/https://arxiv.org/abs/2411.10109
  • Licence: copyrighted (author © under arXiv's non-exclusive distribution licence) — cite + archive + our distillate
  • Distillate: Agents built from a person's own self-reports reproduce that person's survey answers at 83% (interview-grounded) / 82% (survey-grounded) / 86% (both) of the person's two-week test-retest ceiling, versus 74% for demographics-only; a free-text "persona paragraph" scores 0.71, below even the demographics baseline (0.74). Self-report grounding also reduces accuracy disparities across racial and ideological groups. The takeaway the skill leans on: the grounding artifact — the interview transcript — is the product, not a written bio.
  • Check-date: 2026-07-27
  • Reads against: argyle-silicon · wang-flattening · kapania-simulacrum · tjuatja-biases · sharma-sycophancy · park-hai-brief — the first four look opposed and are not, and the distinction is load-bearing: this entry's 83–86% is grounded in the person's own interview, and its own demographics-only arm scores 74%, which is the number the pessimistic four are actually measuring. Read together they say grounding is the variable, not the model
  • Cited-by: audience.md · templates/PERSONA-template.md

park-hai-brief · Park et al., Stanford HAI policy brief ​

ashokkumar-nature · Ashokkumar et al., direction not magnitude ​

  • Citation: Ashokkumar, A., Hewitt, L., Ghezae, I., Willer, R. Nature (advance online publication, 2026-07-08). doi:10.1038/s41586-026-10742-x.
  • Live: https://doi.org/10.1038/s41586-026-10742-x
  • Archive: pending (Save Page Now triggered 2026-07-27; availability: https://archive.org/wayback/available?url=https://doi.org/10.1038/s41586-026-10742-x)
  • Licence: copyrighted (Springer Nature) — cite + archive + our distillate
  • Distillate: Across a large replication set, LLM simulations track the direction of experimental effects at about r≈0.85 while systematically overestimating their magnitude. This is the evidence for the theatre's hardest rule: a synthetic verdict may state direction, never a magnitude (no "23% would churn").
  • Check-date: 2026-07-27
  • Reads against: tjuatja-biases — direction correlates at r≈0.85 while magnitude is overestimated; together they are the whole case for direction, never a magnitude
  • Cited-by: audience.md · skills/advisor/SKILL.md (the second pyramid)

ls-types · Lewis & Sauro, a taxonomy of synthetic users ​

ls-review · Lewis & Sauro, a review of synthetic-user experiments ​

sharma-sycophancy · Sharma et al., sycophancy is trained in ​

  • Citation: Sharma, M., et al. "Towards Understanding Sycophancy in Language Models." ICLR 2024. arXiv:2310.13548 (2023).
  • Live: https://arxiv.org/abs/2310.13548
  • Archive: http://web.archive.org/web/20260725125159/https://arxiv.org/abs/2310.13548
  • Licence: copyrighted (author © under arXiv's non-exclusive distribution licence) — cite + archive + our distillate
  • Distillate: Sycophancy — telling the user what they want to hear — is a trained-in property of RLHF'd assistants, consistent across several models and tasks. A persona built on such a model inherits that compliance, which is why the theatre's calibration layer suppresses sycophancy explicitly (a synthetic respondent is a pleaser unless corrected).
  • Check-date: 2026-07-27
  • Reads against: park-self-reports — a compliance property of the model that any persona inherits, whatever it is grounded in
  • Cited-by: audience.md · skills/advisor/SKILL.md (useful over agreeable)

tjuatja-biases · Tjuatja et al., LLM response biases ≠ human ones ​

  • Citation: Tjuatja, L., et al. "Do LLMs Exhibit Human-like Response Biases? A Case Study in Survey Design." TACL 12 (2024). arXiv:2311.04076 (2023).
  • Live: https://arxiv.org/abs/2311.04076
  • Archive: http://web.archive.org/web/20260116061015/https://arxiv.org/abs/2311.04076
  • Licence: copyrighted (author © under arXiv's non-exclusive distribution licence) — cite + archive + our distillate
  • Distillate: Tests whether LLMs reproduce known human survey response biases (acquiescence, question-order effects) and finds their biases do not reliably mirror human ones — sometimes absent, sometimes inverted. Caveats how far a synthetic survey respondent can stand in for a human one.
  • Check-date: 2026-07-27
  • Reads against: park-self-reports · ashokkumar-nature — response biases do not mirror human ones even where answers correlate, which is why a synthetic rate is refused while a direction is allowed
  • Cited-by: audience.md

argyle-silicon · Argyle et al., silicon sampling and its diversity limits ​

  • Citation: Argyle, L.P., et al. "Out of One, Many: Using Language Models to Simulate Human Samples." Political Analysis (2023). doi:10.1017/pan.2023.2.
  • Live: https://doi.org/10.1017/pan.2023.2
  • Archive: pending (Save Page Now blocked HTTP 523 on 2026-07-27; availability: https://archive.org/wayback/available?url=https://doi.org/10.1017/pan.2023.2)
  • Licence: copyrighted (Cambridge University Press) — cite + archive + our distillate
  • Distillate: Introduces "silicon sampling" — conditioning an LLM on demographic backstories to simulate human survey samples — and shows it can reproduce some subgroup patterns while collapsing within-group diversity. Backs the caution that synthetic samples flatten variety rather than represent it.
  • Check-date: 2026-07-27
  • Reads against: park-self-reports — same subject, different grounding: demographic backstories here, the person's own interview there. Collapsing within-group diversity is what demographics-only produces, and park-self-reports' 74% arm agrees
  • Cited-by: audience.md

wang-flattening · Wang et al., identity flattening ​

  • Citation: Wang, A., Morgenstern, J., Dickerson, J.P. "Large language models that replace human participants can harmfully misportray and flatten identity groups." Nature Machine Intelligence (2025). arXiv:2402.01908.
  • Live: https://arxiv.org/abs/2402.01908
  • Archive: http://web.archive.org/web/20260607174738/https://arxiv.org/abs/2402.01908
  • Licence: copyrighted (author © under arXiv's non-exclusive distribution licence; journal © Springer Nature) — cite + archive + our distillate
  • Distillate: Finds that using LLMs to replace human participants can harmfully misportray and flatten identity groups — reproducing majority stereotypes and erasing within-group variation. This is the direct evidence for the never-assign-a-bias-from-demographics rule: a demographic backstory produces a caricature, not a person.
  • Check-date: 2026-07-27
  • Reads against: park-self-reports — the pessimistic pole, and it is measuring demographic conditioning; not a contradiction of the interview-grounded result
  • Cited-by: audience.md

kapania-simulacrum · Kapania et al., LLMs as qualitative participants ​

  • Citation: Kapania, S., et al. "'Simulacrum of Stories': Examining Large Language Models as Qualitative Research Participants." CHI 2025. arXiv:2409.19430 (2024).
  • Live: https://arxiv.org/abs/2409.19430
  • Archive: http://web.archive.org/web/20260411143601/https://arxiv.org/abs/2409.19430
  • Licence: copyrighted (author © under arXiv's non-exclusive distribution licence) — cite + archive + our distillate
  • Distillate: Treating LLMs as qualitative research participants yields plausible but hollow "simulacra of stories" that miss the lived specificity of real interviews. Marks the boundary of synthetic personas in qualitative work — a supplement, never a replacement for a real transcript.
  • Check-date: 2026-07-27
  • Reads against: park-self-reports — qualitative use specifically. Even interview-grounded, a simulacrum is a supplement; this bounds the optimistic result rather than denying it
  • Cited-by: audience.md · templates/PERSONA-template.md

Cost routing — cheap-first, conditional on a good verifier ​

frugalgpt · Chen, Zaharia & Zou, FrugalGPT ​

  • Citation: Chen, L., Zaharia, M., Zou, J. "FrugalGPT: How to Use Large Language Models While Reducing Cost and Improving Performance." arXiv:2305.05176 (2023).
  • Live: https://arxiv.org/abs/2305.05176
  • Archive: http://web.archive.org/web/20260722202256/https://arxiv.org/abs/2305.05176
  • Licence: copyrighted (author © under arXiv's non-exclusive distribution licence) — cite + archive + our distillate
  • Distillate: A cascade that queries cheaper models first and escalates only on low confidence can match the best single model's accuracy at up to −98% cost. The evidence for cheap-first-then-escalate routing at decomposition.
  • Check-date: 2026-07-27
  • Reads against: not checked
  • Cited-by: cost.md · dispatching.md

routerbench · Hu et al., RouterBench ​

  • Citation: Hu, Q.J., et al. "RouterBench: A Benchmark for Multi-LLM Routing Systems." arXiv:2403.12031 (2024).
  • Live: https://arxiv.org/abs/2403.12031
  • Archive: http://web.archive.org/web/20260606022825/https://arxiv.org/abs/2403.12031
  • Licence: copyrighted (author © under arXiv's non-exclusive distribution licence) — cite + archive + our distillate
  • Distillate: Cascades beat both any individual LLM and a zero-cost router only when the verifier is good — judge error ≤0.1, deteriorating past 0.2. The load-bearing caveat: cheap-first routing is conditional on a good verifier. In the skill, the review gates are that verifier, so the condition is already met — the caveat reads as a strength, not a risk.
  • Check-date: 2026-07-27
  • Reads against: not checked
  • Cited-by: cost.md · dispatching.md

Repository context files ​

agentsmd-eth · Gloaguen et al., do AGENTS.md files help? ​

  • Citation: Gloaguen, T., Mündler, N., Müller, M.N., Raychev, V., Vechev, M. (ETH Zurich). "Evaluating AGENTS.md: Are Repository-Level Context Files Helpful for Coding Agents?" arXiv:2602.11988 (v2, 2026-06-23).
  • Live: https://arxiv.org/abs/2602.11988
  • Archive: http://web.archive.org/web/20260711121106/https://arxiv.org/abs/2602.11988
  • Licence: copyrighted (author © under arXiv's non-exclusive distribution licence) — cite + archive + our distillate
  • Distillate: A coding-agent benchmark finds repository-level context files (./contributing) do not improve task success rates and add roughly +20% inference cost; LLM-generated context files perform slightly worse than none. The evidence behind "curate the shared guide, don't autogenerate it."
  • Check-date: 2026-07-27
  • Reads against: not checked
  • Cited-by: project-layout.md · AGENTS.md

Method provenance and standards (references, not evidence claims) ​

method-provenance · adapted methods, nothing embedded ​

  • Citation: cookiy — user-research-skill (MIT). agentman — "Synthetic Persona Creator" skill (concepts only).
  • Live: https://github.com/cookiy-ai/user-research-skill · https://agentman.ai/agentskills/skill/synthetic-persona-creator (owner-confirmed 2026-07-27; page verified live: three-layer persona architecture — identity foundation · context seeding · response calibration — plus panel distribution)
  • Archive: http://web.archive.org/web/20260411081018/https://github.com/cookiy-ai/user-research-skill
  • Licence: free — cookiy's user-research-skill is MIT (a copy may be carried later; today only the method shape is adapted, nothing embedded). agentman: no licence stated on the page; concepts taken, nothing embedded (no code carried, so no licence obligation).
  • Distillate: Method lineage, not evidence. cookiy's MIT skill supplied the shape of the qualitative-research flows (its qualitative-research-planner → our persona-interview flow, its synthesize-research-report → our QDA step), adapted through the import gate. agentman supplied the calibration and panel concepts behind the persona response-calibration layer. Recorded so every adaptation is auditable and no vendor wrapper is smuggled in.
  • Check-date: 2026-07-27
  • Reads against: not checked
  • Cited-by: TRADEMARKS.md · PATTERNS.md

standards-cluster · named review standards ​