The Second Cycle of Contribution Happens

The project works toward a second cycle of contribution — the conventions survive a first seed-and-run round, attract a second round of contributors (returning or new), and the practice demonstrably continues without dependency on any single author. Reaching the second cycle is the project's core success metric; one-off authoring is not sustainability.

Why It Is Worth Pursuing

A first cycle of contribution proves that a seed corpus can be authored and that the conventions do not immediately break. A second cycle proves something different and harder: that the conventions are legible to authors who were not present at their creation, that curation stays tractable as the graph grows, and that the practice does not collapse when the founding author's attention moves elsewhere. These are the properties that distinguish a shared practice from a single-author garden.

The Founding Conversation names the second-cycle test directly: "The unit of success is whether the second cycle of contribution happens, not whether any particular artifact ships." The reframing is load-bearing for every design choice the project makes. A decision that improves first-cycle artifact quality at the cost of second-cycle legibility fails the test; a decision that sustains second-cycle legibility at some cost to first-cycle elegance passes. The ordering forces the project to ask, at each choice, what makes the second cycle more likely?

The second-cycle frame also distinguishes this project from an MVP. The first cycle is not a minimum viable product shipped to an audience; it is the first iteration of a discipline. Its success is not adoption metrics but whether the discipline proves sustainable when the author stops being the author — when a second contributor arrives and either completes the cycle or reveals that the conventions cannot hold.

Current Gap

No first cycle has completed yet. The seed graph carries contract and decision nodes — the substrate the practice runs on — but no real nodes in the intended knowledge domain. The prototype/nodes/ corpus holds Contracts (9), Decisions (~32 after this commit group), Glosses (6), Convictions (6 after this commit group), Patterns (3), References (1), Observations (2 after this commit group), Aspirations (will be 3 after this commit group). All of those nodes document the convention layer, not the domain.

The bounded-slice selection that would seed a first cycle is an open question. context/Decision Log.md tracks the "Seed corpus domain" Open Question as unresolved — the project has not decided which slice of the intended domain to seed with 15-30 real nodes that exercise the conventions under authoring pressure. Without that decision, the first cycle cannot start, and the second cycle is gated on the first.

No second author exists. The graph has been authored by one person with agent assistance; the conventions have not been read, applied, corrected, or challenged by a contributor who was not present at their creation. Every point at which the conventions might confuse a new reader is currently invisible, because there is no new reader.

The conventions have not been tested under contribution pressure. Authoring under composition-time pressure (what predicate should I use? what shape does this node take?) reveals which conventions are clear, which are under-specified, and which actively obstruct authoring. The current seed graph has been authored deliberately and slowly; a contribution-pressure test has not yet happened.

Work It Asks

Resolve the "Seed corpus domain" Open Question in context/Decision Log.md — decide which bounded slice of the intended domain to seed first. The decision is the immediate blocker on starting the first cycle; without it, every subsequent move is postponed.

Seed 15-30 real nodes in the selected domain slice. The node count is the threshold at which the full predicate family, the form contract catalog, and the curation cadence all get exercised. Fewer than 15 does not stress the conventions; more than 30 risks committing to the conventions at scale before a contribution-pressure test has happened.

Invite 2-4 initial contributors to author in the same graph. The contribution-pressure test is what reveals which conventions hold and which fail. The contributors do not all need to finish nodes; even a single contributor attempting and abandoning a node in the conventions produces high-value information about where the conventions obstruct.

Support a curation ritual cadence through the first cycle. The Pattern Reconcile the Standing Account names the move; the cadence is what turns the move into practice. A weekly or bi-weekly curation pass during the first cycle holds the graph against the write-only drift that Observation Wikis Without Curation Drift Toward Write-Only names as the failure mode this Aspiration targets.

Observe and record what breaks, what holds, what contributors need. Every observation from the first cycle is input to whether the second cycle can be invited. A first cycle that produces only nodes (without recorded observations about the authoring experience) wastes the test.

Progress Recognition

The first bounded slice is selected and seeded with 15-30 nodes that exercise the full predicate family, form contract catalog, and curation cadence.

A second contributor authors their first node in the graph without convergence pressure from the curator — their predicates, their naming, their annotations enter the graph alongside the founder's, not as translations into the founder's vocabulary.

A curation pass reconciles the graph without ontology arbitration. The pass runs; drift is caught; the graph stays legible; no decision gets stalled in "which term is right" arbitration.

A returning or new contributor arrives for the second cycle, reads the existing conventions and the existing nodes, and finds the ground legible enough to begin authoring. The contributor does not need the founder to explain the conventions or the nodes in real-time; the graph's own structure carries enough context to onboard.

The project reaches a steady state where contribution happens on at least two contributors' cadences, curation happens on at least one contributor's cadence, and the practice's continuation does not depend on any single author being present this week.

Sources

Relations