pki.sgit.ai / documents / the-resource-pool

The Resource Pool: A Grant That Depletes, And The First Loss Data This Pivot Can Have

TypeStrategy brief — memo 11, and the series was called complete at ten Versionv0.33.83 Date1 September 2026 AuthorDinis Cruz (project lead) and collaborators LicenceCC BY 4.0 Sourceraw markdown · view on GitHub

Summary

The second axis. Every memo before this one rates CAPABILITY — what an agent can reach. This one rates CONSUMPTION, which is orthogonal to all of it: two placements with byte-identical grant trees can differ by two orders of magnitude in what they burn, and nothing the estate measures would see it. A pool of tokens underwritten for a population rather than allocated per agent, on the reasoning that one execution in a hundred spiking is fine where a hundred spiking is not — which is correlated risk, named in a single line, and the first concrete instance of the aggregation problem doctrine 01 recorded as a rule and never implemented. A resource pool is a grant that DEPLETES: every grant in this corpus is static, and a pool carries a remaining that no existing node has, which makes it a fourth primitive rather than a restatement. The largest thing here is not any of that. The pivot has argued about insurance for eleven memos with no loss data, no claim shape and no pooling, and this supplies all three at once — in a currency that is not money, from meters that already exist for billing, which is why it needs no carrier and puts nobody in the line. A budget overage is a loss event that is dated, quantified, attributable and independently recorded by the supplier's invoice, so loss-event/v0 can finally be drafted against a real instance. It also resolves GM-D78's collision and finds doctrine 07's first mover, because the resource supplier has a reason to refuse that nobody had to argue it into: it is paying. Corrected: the memo says the pool defines the grant, and it does not. A budget bounds volume, never reach, and the cheapest catastrophic action is cheap.

Key concepts

Key ideas

On this site

Adds insurance doctrine 11 and reopens a series called complete at ten. Proposes GM-D86 to GM-D95: consumption as a second axis and the pool as a fourth primitive; a pool bounds volume never reach (correcting the memo); pooling in a non-money currency; consumption as the first obtainable loss data; GM-D78's collision resolved; the supplier as first relying party; the only warranty that cannot be unknown; the excess-of-loss structure and its missing per-occurrence limit; the generalisation's three-part boundary; and a desirable currency making 'never declared, only derived' load-bearing. The v0.33.82 series gate fired on doctrine 10 when this memo was filed, and was rewritten to forbid a hand-typed denominator rather than check one.

Read the document

📄 Original document · v0.33.83 · 1 September 2026 · rendered from the raw markdown (the source of truth)