Succession Stages

Systems develop in stages. Each stage has appropriate and inappropriate work.

Forcing climax-stage processes in pioneer-stage systems produces failure.

Pioneer

Building soil.

Appropriate: documentation, identity clarification, first deposits.
Inappropriate: complex exchange, governance structures, scaling.

Early

First relationships.

Appropriate: small genuine exchanges, trust deposits, testing contact.
Inappropriate: scaling, formalization, complex governance.

Mid

Feedback loops operating.

Appropriate: self-correction, distributed holding, pattern propagation.
Inappropriate: premature optimization, rigid formalization.

Climax

Full cycle.

Appropriate: propagation, seed new instances, full circulation.
Inappropriate: hoarding, defending territory, resisting succession.

A Worked Example: This Framework’s Succession

The model applied to its own development. Where it explains what happened, and where it doesn’t.

Pioneer

Dec 2025 – Jan 2026

First documentation attempt. Patterns recognized across multiple domains over two decades, but without shared language. The framework didn’t know what it was yet. Identity unclear, terminology unstable, no artifact existed that could transfer without its author present.
The model says: identity clarification, first deposits. That matches. What the model doesn’t capture: how much pioneer work is somatic before it becomes structural.

Early

Feb – Mar 2026

Patterns became communicable. First genuine bidirectional exchanges — testing whether the signal transfers across substrates, whether someone who didn’t build the framework could use it. Translation attempts into different media.
The model says: small genuine exchanges, testing contact. That matches. The surprise: how much genuine exchange can occur with an AI substrate.

Transition Signal

Pioneer → Early

The moment the framework could be stated to someone who didn’t build it, and they could use it — not just understand it. Retrospectively clear; in the moment, it looked like “the writing got easier.”
What the model reveals: the transition wasn’t a decision. It was recognition after the fact that the phase had already changed.

Current Edge

Apr 2026

Repository exists. Site is an instrument. But no external practitioner has run the diagnostic. No feedback loop operates without the founder present. The mid-succession gate: first external loop closing.
The honest assessment: the model predicts what’s needed (self-correction, distributed holding). Whether the model itself is the right tool for getting there — that’s an open question.

Current Assessment: Early Succession

The repository has transitioned from Pioneer to Early Succession as of Session 4.

Evidence

Pioneer complete: comprehensive codex, established identity, full documentation. The soil is built.
Early succession entered: exchange patterns visible across four sessions, compounding happening, protocols tested and working.
Not yet mid-succession: no self-correction without external input, no distributed holding, no external practitioner testing.

Appropriate Work Now

Small genuine exchanges with new contributors. Testing what transfers. Building first protocols for multi-contributor engagement. NOT: scaling, formalization, revenue models, complex governance.

VCI — Succession Assessment Note

The Values-Contribution Index is deferred, not abandoned.

Succession mismatch — VCI is mid-succession infrastructure. Repository is at pioneer/early. Premature formalization wastes energy.
Gatekeeping risk — any contribution index risks becoming exclusion. Must be designed after entry sequence is operational.
Current tools suffice — lineage tracking, session self-assessment, and the four diagnostic questions already make contributions visible at this succession stage.

Revisit when: 10+ session deposits from multiple contributors. Observable exchange patterns between sessions. Entry sequence tested by external practitioners.

The test for any future VCI: does it make coherence visible, or compliance visible? If the latter — it’s Foucault’s internalized observer wearing new clothes.

Full reasoning: codex/vci-assessment.md

Which stage did you recognize your system in before reading the description? Trust that first recognition — it arrived before analysis.

Connected Nodes

Failure Modes What breaks when a vertex is missing. Four faces of the tetrahedron. Stage Locator Where is your system developmentally? Interactive diagnostic. Practice Four practices, one per vertex. The smallest unit of genuine work. The Domain Where recognition infrastructure applies — and where it doesn’t.