what we reject
what few explicitly rejects
this document exists to protect the signal and build long-term trust with the 2.5%.
core principle
selection is never for sale. tools earn their place through vetting alone — money cannot buy inclusion, coverage, or a better verdict, and no payment is ever accepted in exchange for a listing. after a tool has independently passed the criteria, clearly-marked affiliate links or sponsorships from vetted sources are acceptable revenue. money may follow the verdict; it may never precede it. other revenue: reader-paid soundings, zaps, premium tools. (revised 2026-07-04: affiliates/sponsors from vetted sources allowed with clear marking.)
what we will never do
- include any tool or project with token premines or fiat-vc dynamics (exit-driven, token games, hype); aligned bitcoin-native capital is judged on the product's merits
- include anything that requires kyc
- accept payment, sponsorship, or favors in exchange for inclusion or a verdict (affiliate/sponsor relationships exist only with already-vetted tools, always clearly marked)
- chase growth, views, or attention
- compromise criteria for popularity or "relevance"
- feature tools connected to ethereum, altcoins, or wrapped assets
- promote hype, marketing campaigns, or kol-driven projects
- change our standards based on market cycles or external pressure
how trust is built
we understand that the 2.5% will not trust us immediately. trust will be earned the same way it is earned with jeff booth — through radical consistency over time.
- we will not change our message or standards
- we will stay small and humble
- we will reject popular tools when they fail the criteria
- we will remain transparent about how we make money
why this matters
in a world full of compromised incentives, the 2.5% need at least one place where the filter is honest. few exists to be that place.
if we ever violate these principles, the project should be abandoned.
the criteria
few tool criteria
strict filter for what qualifies as few-approved. applied through a jeff booth lens.
mandatory requirements
-
bitcoin native
must be built on bitcoin l1 or secured by it. no ethereum derivatives, wrapped tokens, or l2s unless they are genuinely bitcoin-secured and cannot be easily gamed. -
non-custodial
users must retain full control of their keys and funds at all times. -
no premine / no fiat-vc dynamics / no kols
no premined tokens and no paid kol promotion. on capital: fiat-vc dynamics disqualify — exit-driven pressure, token games, hype cycles, growth-at-all-costs. aligned bitcoin-native capital (e.g. ego death capital) does not automatically disqualify; the product is assessed on its own merits against every other criterion. the question is never "who funded it" but "can the incentives be gamed." (clarified 2026-07-04.) -
organic growth
adoption must be genuine and community-driven. no hype marketing or artificial inflation. -
cannot be easily gamed long-term (elevated priority)
this is one of the most important filters. the project must have strong game-theoretic resistance. any system that introduces trusted parties, coordination mechanisms, or incentives that can be exploited over time will be heavily scrutinized or excluded. honest foundations that cannot be cheated long-term are non-negotiable. -
open source or verifiable
code should be open source, or the system must be independently verifiable. -
privacy respecting
default to strong privacy. no unnecessary data collection.
strong preference
- minimal attack surface
- low or no reliance on trusted third parties
- clear alignment with sound money principles
- long-term sustainability over short-term features
note on ecash: excluded from the directory as a wealth-storage recommendation (mints are trusted parties). bounded use as a spending float in few's own services is governed by ecash-position.md.
automatic disqualifiers
- any connection to ethereum or other altcoin ecosystems
- fiat-vc dynamics (exit-driven, token games, hype) or token premines
- heavy kol marketing
- custodial elements
- opaque or unverifiable systems
- high hype / low substance
- introduction of new trusted parties or coordination risks that can be gamed
evaluation process
every tool will be scored on: 1. bitcoin purity 2. sovereignty protection 3. game-theoretic soundness (long-term honesty) 4. practical usefulness to the 2.5%
only tools that score highly across all categories will be included.
position: ecash
few's position on ecash (cashu / fedimint)
written 2026-07-04 to resolve an apparent contradiction: the few directory excludes cashu and fedi over trust assumptions, while the thailand spend service uses ecash mints. the distinction is deliberate, and it comes down to one question: savings or spending?
the principle
few's hardest filter is "cannot be gamed long-term." ecash mints introduce a trusted party — the mint operator can rug or silently inflate. that disqualifies ecash as a place to store wealth, which is what a directory recommendation implies. no amount of ux or privacy changes this: wealth storage with a trusted third party is the exact failure mode bitcoin exists to end.
spending is a different risk profile. money moving through an ecash balance for minutes to days, in amounts you'd carry as street cash, is bounded trust — comparable to the cash in your pocket, which can also be lost or stolen. in exchange, blind signatures buy real privacy at the payment layer that raw lightning doesn't provide. that trade is rational and priced in.
the rules
- ecash is a transient spending float, never a savings vehicle. load just-in-time, spend, withdraw remainder back to lightning/l1 promptly.
- never hold more in a mint than you would carry as cash on the street.
- prefer small, local, auditable mints; spread across mints rather than concentrating.
- the directory does not list ecash mints as recommended tools. the spend service may use them internally, with this position disclosed.
one sentence for the 2.5%
we reject trusted parties for storing your wealth; we accept bounded, cash-like trust for spending it privately — and we tell you exactly where the line is.