the-custodian/workplans/CUST-WP-0050-repo-classification-registration-redesign.md

---
id: CUST-WP-0050
type: workplan
title: "Repo Classification & State Hub Registration Redesign"
domain: custodian
repo: the-custodian
status: active
owner: custodian
topic_slug: custodian
planning_priority: high
planning_order: 50
created: "2026-06-22"
updated: "2026-06-22"
started: "2026-06-22"
state_hub_workstream_id: "9f031f48-8de8-48b6-8e69-d2d83ad70a7a"
---

# CUST-WP-0050 - Repo Classification & State Hub Registration Redesign

## Goal

Adopt the **Repo Classification Standard** (`canon/standards/repo-classification-standard_v1.0.md`,
`id: canon-repo-classification`) as the ecosystem-wide model for organising
repositories, and redesign State Hub registration around it so that:

- every known repository carries a committed `.repo-classification.yaml` that is
  the **source of truth** for its classification,
- the State Hub can **automatically register all known repos** by reading and
  validating those files (local checkout or Gitea API), and
- all **previously registered repos are reclassified** under the new standard,
  replacing the current ad-hoc 14-domain model.

End state: one principled, validated taxonomy (category · domain · capability
tags · business stake · business mechanics) spanning the whole portfolio, with
registration that is reproducible from repo-owned metadata rather than
hand-curated DB rows.

## Context

A 2026-06-21 review compared three views of the portfolio and found them out of
sync:

- **Gitea** hosts ~72 repos (70 under `coulomb/`, plus a fork and a personal repo).
- **State Hub** has 57 `managed_repos` across **14 ad-hoc domains** (custodian,
  railiance, markitect, coulomb_social, personhood, capabilities, canon,
  citation_evidence, helix_forge, inter_hub, netkingdom, stack,
  vergabe_teilnahme, whynot).
- **the-custodian** `canon/projects/` froze at the original **6 founding charters**.

Concrete discrepancies to resolve as part of this work:

- ~18 Gitea repos are **unregistered** (e.g. audit-core, binect-chrome, binect-js,
  coordination-engine, direkt-vermittlung-de, human-resources, polycode-sim,
  ralph-workplan, repo-seed, tegwick-control, tele-mcp, testdrive-jsui,
  timeline-svg, vantage-point, whynot-control, whynot-design).
- **Phantom / renamed** registrations: `markitect-project` (registered) vs
  `markitect-main` (Gitea) — likely a rename; `railiance-bootstrap` and
  `railiance-hosts` registered but absent from Gitea.
- **Duplicate domain**: `vergabe_teilnahme` looks like a second registration of
  `vergabe-teilnahme` (already under coulomb_social).
- **Empty domain**: `personhood` has a charter and topic but no repos.
- **Naming drift**: `coulomb.social`/`coulomb_social`, `foerster-capabilities`/`capabilities`.

The new standard fixes the root cause: it separates *category* (work mode),
*domain* (intended market/user), *capability tags* (what it does), and *business
stake* (who cares) — concerns the current 14 "domains" conflate.

### Architecture decision: repo-anchored model, domain derived from classification

The standard's `domain` is a **fixed 14-value market vocabulary** (infotech,
financials, communication, consumer, health, industrials, energy, utilities,
materials, realestate, crypto, agents, space, government) that is *orthogonal* to
the Hub's current 14 coordination domains. Per the steering decision on
2026-06-22, the new market-domain vocabulary **replaces** the Hub's domain model
(rather than augmenting it or running a parallel two-axis model).

The current spine is `Domain → Topic → Workstream`, where `topics.domain_id` and
`workstreams.topic_id` are both **NOT NULL** and the 14 domains are seeded **1:1
with 14 topics** (a data convention — the schema actually allows many topics per
domain, but that has never been used). `workstreams.repo_id` and
`repo_goals.repo_id` already exist, but the *required* anchor is the soft,
hub-only `topic`, while the stable git-managed `repo` link is optional.

Per the 2026-06-22 steering decision, this redesign **flips the polarity**: the
**repo becomes the primary anchor** for workplans, and market-domain is
**derived** from the repo's `.repo-classification.yaml`, not stored as a separate
`topic`/`domain` parent. Concretely:

- `workstreams.repo_id` becomes the **required** anchor; `topic_id` is demoted to
  optional (or `topic` is retired) — see T10.
- Market-domain is computed from `repo → classification.domain`; the standalone
  `topics.domain_id` / `managed_repos.domain_id` spine is removed.
- `RepoGoal` (already repo-anchored) becomes the goal primitive; `DomainGoal`
  becomes a thin strategic rollup keyed by the 14 market domains.
- **Cross-repo workplans** anchor to a dedicated **project repo** that retires to
  archive on completion, with results living on in the modified product repos —
  see **ADR-005** and Open Questions D1/D1a.

This is consistent with ADR-001: the spine becomes the git-managed repo plus its
committed classification file, so the Hub stays fully rebuildable from repo-owned
files. It is a **breaking migration** of the coordination spine (T04/T05).

## Scope

In scope:

- Promote and steward the standard as custodian canon (done: the standard now
  lives at `canon/standards/repo-classification-standard_v1.0.md`).
- A single machine-readable allowed-values source derived from the standard,
  consumed by both the per-repo files and the Hub validator.
- A committed `.repo-classification.yaml` for every active repo (agent-assisted
  first pass, human-reviewed), authored in each repo.
- State Hub schema/model redesign replacing the domain model with the 14 market
  domains and storing the full classification on `managed_repos`.
- A reversible data migration re-homing existing topics/workstreams/goals/
  decisions/charters and resolving the discrepancies listed above.
- Auto-registration tooling (bulk, idempotent) that reads classification files
  from local checkouts or the Gitea API and registers/reclassifies repos.
- Updates to dashboard, consistency checker, MCP/REST surface, and orientation
  docs to the new taxonomy.

In scope (added 2026-06-22):

- Re-anchor workplans to repos (`repo_id` required, `topic` optional/retired) and
  derive market-domain from classification (T04/T10).
- Rename `workstream → workplan` across schema, API, and MCP so the Hub
  vocabulary matches the repo files and current usage (T10).

Out of scope:

- Re-architecting task-level semantics beyond what the re-anchor and rename force.
- Changing the Gitea hosting model or repo contents beyond adding the
  classification file (and, for cross-repo efforts, creating project repos per
  ADR-005).
- Classifying throwaway/forked/non-ecosystem repos (explicit exclusion list).

## Repo boundary

This is the **custodian driving/coordination workplan** (it owns the canon
standard and the portfolio decision), consistent with how `CUST-WP-0043` drove
State Hub work. Implementation tasks **T04–T08 execute in `/home/worsch/state-hub`**
and should be re-homed as a state-hub-local workplan once this plan is approved;
per-repo classification files (T02/T03) are authored in each target repo. The
hub remains a read/index model fed by repo-owned files (ADR-001).

## Tasks

### Phase 1 — Standard as a validation source

### T01 - Derive machine-readable allowed-values from the standard

```task
id: CUST-WP-0050-T01
status: done
priority: high
state_hub_task_id: "d978b1f3-4eca-4a17-835b-2c25d13cae22"
```

Extract the standard's controlled vocabularies (5 categories, 14 domains, the
business_stake and business_mechanics enums, and the recommended capability
families) into a single machine-readable artefact (e.g.
`canon/standards/repo-classification.allowed.yaml`) that both the per-repo
`.repo-classification.yaml` linter and the State Hub validator import.

Done when a single allowed-values file exists, is referenced by the standard, and
a small validator can check a `.repo-classification.yaml` against it.

**Delivered (2026-06-22):** `canon/standards/repo-classification.allowed.yaml`
(categories, domains, business_stake, business_mechanics, capability families,
guidance bounds); referenced from the standard §12; validator
`tools/validate_repo_classification.py` (stdlib + PyYAML) with `--self-test`
(PASS) — checks category/domain enums, secondary-domain rules, kebab-case tags,
and stake/mechanics enums.

### Phase 2 — Classify the portfolio (repo-owned source of truth)

### T02 - Classify custodian-owned repos

```task
id: CUST-WP-0050-T02
status: done
priority: high
state_hub_task_id: "b7edfbb5-483f-4600-9356-8f885c78ce58"
```

Author and human-review `.repo-classification.yaml` for the custodian-domain
repos (the-custodian, state-hub, hub-core, inter-hub, activity-core, issue-core,
kaizen-agentic, llm-connect, ops-bridge, ops-warden, email-connect) using the
standard's §16 agent prompt as a first pass.

Done when each custodian repo has a committed file that validates against T01 and
has been reviewed by a human.

**Progress (2026-06-22):** all 11 custodian-domain repos now carry a committed,
validated `.repo-classification.yaml` (first-pass `classified_by: agent`).
Following the 2026-06-22 decision, a new **`tooling`** category (between `project`
and `product`) was added to the standard for reusable internal
tooling/infrastructure, and the nine tooling repos were reclassified to it:
the-custodian (research·infotech), inter-hub (research·infotech), state-hub
(tooling·infotech), hub-core (tooling·infotech), activity-core (tooling·infotech),
issue-core (tooling·infotech), kaizen-agentic (tooling·agents), llm-connect
(tooling·agents), ops-bridge (tooling·infotech), ops-warden (tooling·infotech),
email-connect (tooling·infotech). Commits are **local-only** in each repo (not
yet pushed to Gitea).

**Done (2026-06-22):** human review complete — Bernd confirmed the
agents-vs-infotech primary-domain choice, keeping both kaizen-agentic and
llm-connect as `agents` primary (`infotech` secondary). All 11 files flipped to
`classified_by: human` and re-validated clean against T01. Task **done**.

### T03 - Classify the full Gitea inventory — DROPPED

```task
id: CUST-WP-0050-T03
status: cancel
priority: high
state_hub_task_id: "81489716-61ef-4207-ab8a-5877843281de"
```

**Dropped 2026-06-22.** T03 conflated *authoring classification files* with
*registering repos*. Doing a ~70-repo PR storm before the vocabulary and schema
are proven is premature and high-blast-radius, and registering the ~18
unregistered repos now would land them in the **old** domain model — creating
legacy only to clean it up. Classifying + registering the remaining inventory is
deferred to **T11**, executed **under the new model after cutover**. The 11
custodian fixtures (T02) plus T01 are sufficient to build and prove the redesign.

### Phase 3–4 — RE-HOMED to STATE-WP-0065 (state-hub)

Per this plan's repo boundary and the 2026-06-22 decision, the implementation of
the State Hub redesign now lives in a **state-hub-local workplan**:
`state-hub/workplans/STATE-WP-0065-repo-anchored-classification-spine.md`
(workstream `8dc7d106-11e2-41df-b512-89ed69d2a65f`). CUST-WP-0050 remains the
**coordination driver** (canon standard, decisions D1/D1a, ADR-005). The original
implementation tasks below are **cancelled here** (re-homed, not abandoned); the
efficient regrouping merges the three spine-rewriting tasks into one migration:

| Was (CUST-WP-0050) | Now (STATE-WP-0065) |
| --- | --- |
| T04 schema + T05 data migration + T10 re-anchor/rename | **P1** single Alembic spine migration |
| T04/T10 API + validation surface | **P2** API / MCP / validation |
| T06 auto-registration | **P3** auto-registration tooling |
| T07 reclassify existing | folded into **P3** (lazy, as committed files appear) |
| T08 surfaces + T09 cutover | **P4** surfaces & cutover |

```task
id: CUST-WP-0050-T04
status: cancel
priority: high
state_hub_task_id: "b61f6267-c2b2-4325-95fa-30ee899ce7d1"
```
Re-homed → STATE-WP-0065 P1 (schema: domains→14 market domains + classification on `managed_repos`).

```task
id: CUST-WP-0050-T05
status: cancel
priority: high
state_hub_task_id: "171fa385-4d78-41ea-b749-ac3f9082fe47"
```
Re-homed → STATE-WP-0065 P1 (data migration + discrepancy resolution, same window as schema).

```task
id: CUST-WP-0050-T06
status: cancel
priority: high
state_hub_task_id: "6ae14007-d6d2-4395-814e-ace91486a953"
```
Re-homed → STATE-WP-0065 P3 (`register-from-classification` bulk/idempotent tooling).

```task
id: CUST-WP-0050-T07
status: cancel
priority: medium
state_hub_task_id: "6411bf3f-9de2-4bcd-9ffe-6209cda6ba93"
```
Re-homed → STATE-WP-0065 P3 (lazy reclassification of existing registrations as committed files appear).

```task
id: CUST-WP-0050-T08
status: cancel
priority: medium
state_hub_task_id: "09951aec-2960-4c50-b73d-4e2e7bd285c9"
```
Re-homed → STATE-WP-0065 P4 (dashboard, consistency rule, MCP/REST filters, docs).

```task
id: CUST-WP-0050-T09
status: cancel
priority: medium
state_hub_task_id: "babbb80a-c52d-4ec2-b217-2f6196a2e5f3"
```
Re-homed → STATE-WP-0065 P4 (cutover, verification, retire old model).

```task
id: CUST-WP-0050-T10
status: cancel
priority: high
state_hub_task_id: "bee16416-a67f-4155-93d7-09f278daa04f"
```
Re-homed → STATE-WP-0065 P1 (re-anchor `repo_id` required + `workstream → workplan` rename, merged into the spine migration).

### Phase 4 (custodian) — Post-cutover inventory

### T11 - Classify & register remaining Gitea inventory (post-cutover)

```task
id: CUST-WP-0050-T11
status: todo
priority: medium
```

Replaces dropped T03. **After** STATE-WP-0065 cutover proves the new model,
author `.repo-classification.yaml` for the remaining active Gitea repos (the ~18
unregistered + any not yet migrated) and bulk-register them via the
STATE-WP-0065 P3 tooling — **under the new model**, no legacy detour. Maintain an
explicit **exclusion list** (fork `tegwick/the-custodian`,
`lando_worsch/python-snake`, archived `test_domain_v2`, inactive repos).

Done when every non-excluded active Gitea repo has a committed, validated
classification file and a `managed_repo` row under the new taxonomy (or is on the
recorded exclusion list).

## Open Questions / Decisions

- **D1 (RESOLVED 2026-06-22): the repo is the primary anchor.** Workplans bind to
  repos (`repo_id` required); market-domain is *derived* from the repo's
  classification; `topic`/`domain` stop being the spine (`topic` retires or
  becomes an optional cross-repo tag). This supersedes the earlier "keep topic as
  an independent coordination unit" proposal. Implemented by T04/T10.
- **D1a (open, follows from D1): anchor for cross-repo workplans.** Per **ADR-005**,
  a complex cross-repo effort gets its own **project repo** (`category: project`)
  as the anchor, retired to archive on completion with results living in the
  modified product repos. Open sub-point: the project-repo **naming convention**
  (e.g. `proj-<slug>` vs a dedicated grouping) and the archival trigger details.
- **D2: classification ownership/approval.** Who approves each repo's
  `.repo-classification.yaml` — per-repo owner, or central custodian review?
- **D3: exclusion list.** Confirm exclusions (fork `tegwick/the-custodian`,
  `lando_worsch/python-snake`, archived `test_domain_v2`, any inactive repos).
- **D4: behavioural vs descriptive.** Do `secondary_domains` / `capability_tags`
  / `business_stake` drive any Hub behaviour initially, or are they descriptive
  until a later phase?

## Risks

- **Breaking-migration blast radius** — topics/workstreams/goals/decisions and
  charter `topic_id` references all move; mitigate with a reviewed dry-run and a
  tested downgrade (T05).
- **Cross-repo coordination** — T03 touches ~70 repos via PRs; sequence behind
  T01/T02 so the vocabulary is stable first.
- **Consistency-checker coupling** — existing C-rules assume the current domain
  model; update alongside (T08) to avoid mass false positives.
- **Boundary drift** — keep implementation in `state-hub`; this plan coordinates.