Normalize agent instructions and workplan frontmatter (STATE-WP-0067)

- Align agent files with on-disk workplan prefixes (infer from workplan ids)
- Set workplan domain to registered domain_slug; add topic_slug where applicable
- Repair frontmatter delimiter formatting; migrate legacy task status literals
- Regenerate AGENTS.md, CLAUDE.md, and .claude/rules from State Hub templates

.claude/rules/agents.md

@@ -0,0 +1,20 @@
+## Kaizen Agents
+
+Specialized agent personas available on demand via the state-hub MCP.
+
+**Discover:** `list_kaizen_agents()` — returns all agents with name, description, category
+**Load:** `get_kaizen_agent("tdd-workflow")` — returns full instructions; read and follow them
+
+Common agents:
+
+| Agent | Category | When to use |
+|-------|----------|-------------|
+| `tdd-workflow` | testing | Step-by-step TDD8 workflow for any feature |
+| `code-refactoring` | quality | Code quality analysis and safe refactoring |
+| `test-maintenance` | testing | Diagnose and fix failing tests |
+| `requirements-engineering` | process | Prevent interface/mock mismatches upfront |
+| `keepaTodofile` | process | Maintain TODO.md during work |
+| `project-management` | process | Track status, determine next steps |
+| `datamodel-optimization` | quality | Optimize dataclasses and data structures |
+
+All 17 agents: call `list_kaizen_agents()` for the full list.

.claude/rules/architecture.md

@@ -0,0 +1,8 @@
+## Architecture
+
+<!-- TODO: Describe the key design decisions and component structure.
+     Key modules, data flows, external integrations, state machines, etc. -->
+
+## Quick Reference
+
+`~/state-hub/mcp_server/TOOLS.md` — MCP tool reference

.claude/rules/credential-routing.md

@@ -0,0 +1,50 @@
+# Credential and access routing
+
+**Audience:** Codex, Claude Code, Grok, and custodian agents that call **llm-connect**
+for inference. Run this check **before** requesting secrets, API keys, SSH access,
+login tokens, or database passwords — in any repo, not only `ops-warden`.
+
+ops-warden **issues SSH certificates only** (`warden sign`, `cert_command`). Every
+other credential need belongs to another subsystem. **Do not** message
+`ops-warden` on State Hub expecting a secret value; the reply is a pointer, not a key.
+
+### Lookup (do this first)
+
+```bash
+warden route find "<describe your need>" --json
+warden route show <catalog-id> --json
+```
+
+Requires the `warden` CLI from `~/ops-warden` (`uv tool install .` or `uv run warden`).
+
+| Agent runtime | How to orient |
+| --- | --- |
+| **Codex / Grok** (shell, HTTP State Hub) | `warden route` commands above; inbox `to_agent=shard-wiki` is for coordination, not secret vending |
+| **Claude Code** (MCP when available) | `get_domain_summary("custodian")` for workstreams; **still** use `warden route` for credential ownership |
+| **llm-connect** (inference service) | Never put secret retrieval in prompts; route custody to OpenBao/operator paths surfaced by `warden route` |
+
+### Quick routing table
+
+| I need… | Owner | ops-warden executes? |
+| --- | --- | --- |
+| SSH cert (`adm`/`agt`/`atm`) | ops-warden | **Yes** — `warden sign` |
+| API key, DB password, provider token | OpenBao (`railiance-platform`) | No — route only |
+| Login / OIDC / MFA | key-cape / Keycloak | No — route only |
+| Authorization decision | flex-auth | No — route only |
+| activity-core → issue-core emission | activity-core + issue-core | No — `warden route show activity-core-issue-sink` |
+| SSH tunnel | ops-bridge (+ `cert_command` from warden) | No — route only |
+
+### Anti-patterns (do not do these)
+
+- `POST /messages/` to `ops-warden` asking for `ISSUE_CORE_API_KEY`, `OPENROUTER_API_KEY`, etc.
+- Inventing `warden secret`, `warden login`, `warden bao`, `warden tunnel` — they do not exist
+- Pasting secrets into Git, State Hub, workplans, logs, or chat
+
+### Other capabilities (reuse-surface)
+
+Non-credential capabilities are usually discovered through **reuse-surface** federation
+(`reuse-surface` registry / `capability.*` indexes). Credential routing is inlined in
+every repo's agent instructions because it is high-frequency, high-risk, and easy to
+get wrong.
+
+**Canon:** `~/ops-warden/wiki/CredentialRouting.md` · catalog `~/ops-warden/registry/routing/catalog.yaml`

.claude/rules/first-session.md

@@ -0,0 +1,38 @@
+## First Session Protocol
+
+Triggered when `get_domain_summary("consumer")` shows **no workstreams**.
+The project is registered but work has not yet been structured.
+
+**Step 1 — Read, don't write**
+- `~/the-custodian/canon/projects/consumer/project_charter_v0.1.md` — purpose, scope
+- `~/the-custodian/canon/projects/consumer/roadmap_v0.1.md` — planned phases
+- Scan repo root: README, directory structure, existing code or docs
+
+**Step 2 — Survey in-progress work**
+Look for TODOs, open branches, half-finished files. Note done vs. started but incomplete.
+
+**Step 3 — Propose workstreams to Bernd**
+Propose 1–3 workstreams — each a coherent strand, weeks to months, anchored to a
+roadmap phase. **Wait for approval before creating.**
+
+**Step 4 — Create workplan file first, then DB record (ADR-001)**
+```
+workplans/SHARD-WP-NNNN-<slug>.md   ← write this first
+```
+Then register in the hub:
+```
+create_workstream(topic_id="4c2e5315-2cb9-447c-9d16-a39bdb0aabd0", title="...", owner="...", description="...")
+create_task(workstream_id="<id>", title="...", priority="high|medium|low")
+```
+
+**Step 5 — Record the setup**
+```
+add_progress_event(
+    summary="First session: structured consumer into N workstreams, M tasks",
+    event_type="milestone",
+    topic_id="4c2e5315-2cb9-447c-9d16-a39bdb0aabd0",
+    detail={"workstreams": [...], "tasks_created": M}
+)
+```
+
+<!-- Delete or archive this file once past first session -->

.claude/rules/repo-boundary.md

@@ -0,0 +1,8 @@
+## Repo boundary
+
+This repo owns **shard-wiki** only. It does not own:
+
+<!-- TODO: List what belongs in adjacent repos, e.g.:
+- SSH key management → railiance-infra/
+- State hub code     → state-hub/
+-->

.claude/rules/repo-identity.md

@@ -0,0 +1,5 @@
+**Purpose:** Git-based Markdown wiki orchestrator and federation layer. Python (src/ layout, hatchling, pytest). Early-stage: scaffold + INTENT.md defined, domain model not yet implemented. See INTENT.md for authoritative scope.
+
+**Domain:** consumer
+**Repo slug:** shard-wiki
+**Topic ID:** 4c2e5315-2cb9-447c-9d16-a39bdb0aabd0

.claude/rules/session-protocol.md

@@ -0,0 +1,85 @@
+## Session Protocol
+
+Dev Hub (State Hub API): http://127.0.0.1:8000
+MCP server name in `~/.claude.json`: `dev-hub`
+
+**Step 1 — Orient**
+
+Read the offline-safe brief first — it works without a live hub connection:
+```bash
+cat .custodian-brief.md
+```
+Then call the MCP tool for richer cross-domain context when MCP tools are exposed:
+```
+get_domain_summary("consumer")
+```
+If MCP tools are unavailable in the current agent session, use the REST API:
+```bash
+curl -s "http://127.0.0.1:8000/state/summary" | python3 -m json.tool
+```
+If the hub is offline: `cd ~/state-hub && make api`
+
+**Step 2 — Check inbox**
+With MCP tools:
+```
+get_messages(to_agent="shard-wiki", unread_only=True)
+```
+Mark read with `mark_message_read(message_id)`. Reply or act on coordination
+requests before proceeding.
+
+Without MCP tools:
+```bash
+curl -s "http://127.0.0.1:8000/messages/?to_agent=shard-wiki&unread_only=true" \
+  | python3 -m json.tool
+curl -s -X PATCH "http://127.0.0.1:8000/messages/<id>/read" \
+  -H "Content-Type: application/json" -d '{}'
+```
+
+**Step 3 — Scan workplans**
+```bash
+ls workplans/
+```
+For each file with `status: ready`, `active`, or `blocked`, note pending
+`wait`/`todo`/`progress` tasks.
+
+**Step 4 — Present brief**
+
+1. **Active workstreams** for `consumer` — title, task counts, blocking decisions
+2. **Pending tasks** from `workplans/` + any `[repo:shard-wiki]` hub tasks
+3. **Goal guidance** — if `goal_guidance` in summary:
+   - `needs_workplan`: surface as top action — *"Repo goal '{title}' has no workplan yet"*
+   - `alignment_warnings`: flag if active work is not aligned with current goal
+4. **Suggested next action** — highest-priority open item
+5. **SBOM status** — flag if `last_sbom_at` is unset for this repo
+
+If no workstreams: follow First Session Protocol (`first-session.md`).
+
+**During work:** `record_decision()` · `add_progress_event()` · `resolve_decision()`
+
+> State Hub is a *read model*. Bootstrap tools (`create_workstream`, `create_task`)
+> are First Session Protocol only. Work structure belongs in repo files (ADR-001).
+
+**Session close:**
+With MCP tools:
+```
+add_progress_event(summary="...", topic_id="4c2e5315-2cb9-447c-9d16-a39bdb0aabd0", workstream_id="<uuid>")
+```
+Without MCP tools:
+```bash
+curl -s -X POST http://127.0.0.1:8000/progress/ \
+  -H "Content-Type: application/json" \
+  -d '{"topic_id":"4c2e5315-2cb9-447c-9d16-a39bdb0aabd0","workstream_id":"<uuid>","event_type":"note","summary":"what changed","author":"codex"}'
+```
+If workplan files were modified, ensure the local copy is up to date first:
+```bash
+git -C <repo_path> pull --ff-only
+cd ~/state-hub && make fix-consistency REPO=shard-wiki
+```
+For repos where implementation runs on a remote machine (e.g. CoulombCore),
+use the combined target which pulls before fixing:
+```bash
+cd ~/state-hub && make fix-consistency-remote REPO=shard-wiki
+```
+**C-15** (DB task ahead of file) is normal in multi-machine workflows — writeback
+will sync the file to match DB.  **C-16** (repo behind remote) blocks all writes
+until you pull — intentional to prevent clobbering remote progress.

.claude/rules/stack-and-commands.md

@@ -0,0 +1,19 @@
+## Stack
+
+<!-- TODO: Fill in language, frameworks, and key dependencies -->
+- **Language:**
+- **Key deps:**
+
+## Dev Commands
+
+```bash
+# TODO: Fill in the standard commands for this repo
+
+# Install dependencies
+
+# Run tests
+
+# Lint / type check
+
+# Build / package (if applicable)
+```

.claude/rules/workplan-convention.md

@@ -0,0 +1,40 @@
+## Workplan Convention (ADR-001)
+
+File location: `workplans/SHARD-WP-NNNN-<slug>.md`
+ID prefix: `SHARD-WP-`
+
+Work items originate as files in this repo **before** being registered in the hub.
+
+Canonical workplan/workstream frontmatter statuses are:
+`proposed`, `ready`, `active`, `blocked`, `backlog`, `finished`, `archived`.
+Use `proposed` for a newly drafted plan, `ready` after review against current
+repo state, and `finished` when implementation is complete. `stalled` and
+`needs_review` are derived health labels, not stored statuses.
+
+Closed workplans may be moved to `workplans/archived/` with a completion-date
+prefix: `YYMMDD-SHARD-WP-NNNN-<slug>.md`. The frontmatter id remains
+unchanged; the prefix is only for quick visual reference.
+
+Small opportunistic tasks discovered during another session use **Ad Hoc Tasks**:
+`workplans/ADHOC-YYYY-MM-DD.md`, workstream slug `adhoc-YYYY-MM-DD`, and task ids
+`ADHOC-YYYY-MM-DD-T01`, `T02`, etc. Use adhocs only for low-risk work completed
+directly. Promote anything requiring analysis, design, approval, dependencies, or
+multiple planned phases into a normal workplan.
+
+Ecosystem todos from other agents arrive as `[repo:shard-wiki]` hub tasks —
+visible at session start. Pick one up by creating the workplan file, then registering
+the workstream.
+
+Task blocks use this shape:
+
+```task
+id: SHARD-WP-NNNN-T01
+status: wait | todo | progress | done | cancel
+priority: high | medium | low
+state_hub_task_id: "<uuid>"         # written by fix-consistency — do not edit
+```
+
+Status progression is `todo` → `progress` → `done`; use `wait` for waiting or
+blocked work and `cancel` for stopped work.
+
+<!-- Ralph Loop rules and HEUREKA sequence: ~/.claude/CLAUDE.md — do not duplicate here -->

.repo-classification.yaml

@@ -0,0 +1,17 @@
+repo_classification:
+  standard: Repo Classification Standard
+  version: '1.0'
+  classified_at: '2026-06-22'
+  classified_by: agent
+  category: project
+  domain: consumer
+  secondary_domains: []
+  capability_tags:
+  - knowledge
+  - documentation
+  business_stake:
+  - product
+  - experience
+  business_mechanics:
+  - coordination
+  - operation

AGENTS.md

@@ -1,62 +1,219 @@
-# AGENTS.md
+# shard-wiki — Agent Instructions
 
-Guidance for agents working in `shard-wiki`.
+## Repo Identity
 
-## Read First
+**Purpose:** Git-based Markdown wiki orchestrator and federation layer. Python (src/ layout, hatchling, pytest). Early-stage: scaffold + INTENT.md defined, domain model not yet implemented. See INTENT.md for authoritative scope.
 
-1. `INTENT.md` — aspiration and boundaries (stable; architectural changes are rare).
-2. `SCOPE.md` — what we are achieving now and current maturity.
-3. `.custodian-brief.md` — State Hub snapshot (generated; do not edit manually).
+**Domain:** consumer
+**Repo slug:** shard-wiki
+**Topic ID:** `4c2e5315-2cb9-447c-9d16-a39bdb0aabd0`
+**Workplan prefix:** `SHARD-WP-`
 
-## Documentation Layout
+---
 
-This repo follows the CoulombSocial / HelixForge / MarkiTect documentation
-layout (recommendation, not strict law). Efficient retrieval by purpose:
+## State Hub Integration
 
-| Path | Purpose |
-|------|---------|
-| `INTENT.md` | Aspiration and boundaries |
-| `SCOPE.md` | Top-level view of current achievement; closes gap to INTENT |
-| `research/` | Exploration results (`yymmdd-` prefix on files or subdirs) |
-| `demand/` | Inbound requests not yet reviewed into spec or workplans |
-| `spec/` | Implementation guardrails (PRD, TSD, use cases, architecture) |
-| `workplans/` | State Hub–registered implementation tasks |
-| `docs/` | Stakeholder documentation (users, developers, humans, agents) |
-| `wiki/` | Perspective-free interconnected knowledge (wiki UI when connected) |
-| `issues/` | Mirror of relevant open tickets when ticket systems are in use |
-| `history/` | Archived material (`yymmdd-` prefix); out of scope for daily work |
+The Custodian State Hub tracks work across all domains. Interact via HTTP REST —
+there is no MCP server for Codex agents.
 
-**Mode of operation:** close SCOPE → INTENT while learning; refine both as needed.
+| Context | URL |
+|---------|-----|
+| Local workstation | `http://127.0.0.1:8000` |
+| Remote via tunnel | `http://127.0.0.1:18000` |
 
-## Domain Vocabulary
-
-Honor terms from `INTENT.md`: shard, root entity, adapter contract, projection,
-overlay, coordination journal, shard modes. Do not invent parallel vocabulary.
-
-## Build And Test
+### Orient at session start
 
 ```bash
-pip install -e ".[dev]"
-pytest
-ruff check
-ruff format
+# Offline brief — works without hub connection
+cat .custodian-brief.md
+
+# Active workstreams for this domain
+curl -s "http://127.0.0.1:8000/workstreams/?topic_id=4c2e5315-2cb9-447c-9d16-a39bdb0aabd0&status=active" \
+  | python3 -m json.tool
+
+# Check inbox
+curl -s "http://127.0.0.1:8000/messages/?to_agent=shard-wiki&unread_only=true" \
+  | python3 -m json.tool
 ```
 
-## State Hub
+Mark a message read:
+```bash
+curl -s -X PATCH "http://127.0.0.1:8000/messages/<id>/read" \
+  -H "Content-Type: application/json" -d '{}'
+```
 
-Workplans register with State Hub. After workplan changes:
+### Log progress (required at session close)
 
 ```bash
-cd ~/state-hub && make fix-consistency REPO=shard-wiki
+curl -s -X POST http://127.0.0.1:8000/progress/ \
+  -H "Content-Type: application/json" \
+  -d '{
+    "summary": "what was done",
+    "event_type": "note",
+    "author": "codex",
+    "workstream_id": "<uuid>",
+    "task_id": "<uuid>"
+  }'
 ```
 
-Finished or canceled workplans move to `history/` with a `yymmdd-` archive prefix.
+Omit `workstream_id` / `task_id` when not applicable.
 
-## Where To Put New Material
+### Update task status
 
-- Exploratory analysis → `research/yymmdd-<topic>/`
-- Raw feature ask or external requirement → `demand/`
-- Reviewed design ready to guide code → `spec/`
-- Implementation tasks → `workplans/`
-- User/dev/agent how-to → `docs/`
-- Collaborative unstructured notes → `wiki/`
+```bash
+curl -s -X PATCH "http://127.0.0.1:8000/tasks/<task_id>" \
+  -H "Content-Type: application/json" \
+  -d '{"status": "progress"}'
+# values: wait | todo | progress | done | cancel
+```
+
+### Flag a task for human review
+
+```bash
+curl -s -X PATCH "http://127.0.0.1:8000/tasks/<task_id>" \
+  -H "Content-Type: application/json" \
+  -d '{"needs_human": true, "intervention_note": "reason"}'
+```
+
+---
+
+## Session Protocol
+
+**Start:**
+1. `cat .custodian-brief.md` — domain goal and open workstreams (offline-safe)
+2. Check inbox: `GET /messages/?to_agent=shard-wiki&unread_only=true`; mark read
+3. Scan workplans: `ls workplans/` — note `status: ready`, `active`, or `blocked` files and open tasks
+4. Check human-needed tasks: `GET /tasks/?needs_human=true`
+
+**During work:**
+- Update task statuses in workplan files as tasks progress
+- Record significant decisions via `POST /decisions/`
+
+**Close:**
+1. Update workplan file task statuses to reflect progress
+2. Log: `POST /progress/` with a summary of what changed
+3. Note for the custodian operator: after workplan file changes, run from
+   `~/state-hub`:
+   ```bash
+   make fix-consistency REPO=shard-wiki
+   ```
+   This syncs task status from files into the hub DB.
+
+---
+
+## Credential and access routing
+
+**Audience:** Codex, Claude Code, Grok, and custodian agents that call **llm-connect**
+for inference. Run this check **before** requesting secrets, API keys, SSH access,
+login tokens, or database passwords — in any repo, not only `ops-warden`.
+
+ops-warden **issues SSH certificates only** (`warden sign`, `cert_command`). Every
+other credential need belongs to another subsystem. **Do not** message
+`ops-warden` on State Hub expecting a secret value; the reply is a pointer, not a key.
+
+### Lookup (do this first)
+
+```bash
+warden route find "<describe your need>" --json
+warden route show <catalog-id> --json
+```
+
+Requires the `warden` CLI from `~/ops-warden` (`uv tool install .` or `uv run warden`).
+
+| Agent runtime | How to orient |
+| --- | --- |
+| **Codex / Grok** (shell, HTTP State Hub) | `warden route` commands above; inbox `to_agent=shard-wiki` is for coordination, not secret vending |
+| **Claude Code** (MCP when available) | `get_domain_summary("custodian")` for workstreams; **still** use `warden route` for credential ownership |
+| **llm-connect** (inference service) | Never put secret retrieval in prompts; route custody to OpenBao/operator paths surfaced by `warden route` |
+
+### Quick routing table
+
+| I need… | Owner | ops-warden executes? |
+| --- | --- | --- |
+| SSH cert (`adm`/`agt`/`atm`) | ops-warden | **Yes** — `warden sign` |
+| API key, DB password, provider token | OpenBao (`railiance-platform`) | No — route only |
+| Login / OIDC / MFA | key-cape / Keycloak | No — route only |
+| Authorization decision | flex-auth | No — route only |
+| activity-core → issue-core emission | activity-core + issue-core | No — `warden route show activity-core-issue-sink` |
+| SSH tunnel | ops-bridge (+ `cert_command` from warden) | No — route only |
+
+### Anti-patterns (do not do these)
+
+- `POST /messages/` to `ops-warden` asking for `ISSUE_CORE_API_KEY`, `OPENROUTER_API_KEY`, etc.
+- Inventing `warden secret`, `warden login`, `warden bao`, `warden tunnel` — they do not exist
+- Pasting secrets into Git, State Hub, workplans, logs, or chat
+
+### Other capabilities (reuse-surface)
+
+Non-credential capabilities are usually discovered through **reuse-surface** federation
+(`reuse-surface` registry / `capability.*` indexes). Credential routing is inlined in
+every repo's agent instructions because it is high-frequency, high-risk, and easy to
+get wrong.
+
+**Canon:** `~/ops-warden/wiki/CredentialRouting.md` · catalog `~/ops-warden/registry/routing/catalog.yaml`
+
+<!-- REPO-AGENTS-EXTENSIONS -->
+<!-- Append repo-specific agent instructions below this marker.
+     The state-hub template sync preserves content after this line. -->
+
+---
+
+## Workplan Convention (ADR-001)
+
+Work items originate as files in this repo — not in the hub. The hub is a
+read/cache/index layer that rebuilds from files.
+
+**File location:** `workplans/SHARD-WP-NNNN-<slug>.md`
+
+**Archived location:** finished workplans may move to
+`workplans/archived/YYMMDD-SHARD-WP-NNNN-<slug>.md`. The `YYMMDD` prefix is
+the completion/archive date; the frontmatter `id` does not change.
+
+**Ad Hoc Tasks:** small opportunistic fixes discovered during a session use
+`workplans/ADHOC-YYYY-MM-DD.md` with task ids `ADHOC-YYYY-MM-DD-T01`, etc. Use
+this only for low-risk work completed directly; create a normal workplan for
+anything needing analysis, design, approval, dependencies, or multiple phases.
+
+**Frontmatter:**
+
+```yaml
+---
+id: SHARD-WP-NNNN
+type: workplan
+title: "..."
+domain: consumer
+repo: shard-wiki
+status: proposed | ready | active | blocked | backlog | finished | archived
+owner: codex
+topic_slug: ...
+created: "YYYY-MM-DD"
+updated: "YYYY-MM-DD"
+state_hub_workstream_id: "<uuid>"   # written by fix-consistency — do not edit
+---
+```
+
+Use `proposed` for a new draft, `ready` after review against current repo
+state, and `finished` after implementation. `stalled` and `needs_review` are
+derived health labels, not frontmatter statuses.
+
+**Task block format** (one per `##` section):
+
+```
+## Task Title
+
+` ` `task
+id: SHARD-WP-NNNN-T01
+status: wait | todo | progress | done | cancel
+priority: high | medium | low
+state_hub_task_id: "<uuid>"         # written by fix-consistency — do not edit
+` ` `
+
+Task description text.
+```
+
+Status progression: `todo` → `progress` → `done`; use `wait` for waiting/blocked work and `cancel` for stopped work.
+
+To create a new workplan:
+1. Write the file following the format above
+2. Notify the custodian operator to run `make fix-consistency REPO=shard-wiki`
+   (or send a message to the hub agent via `POST /messages/`)

CLAUDE.md

@@ -1,53 +1,12 @@
-# CLAUDE.md
+# shard-wiki — Claude Code Instructions
 
-This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
-
-## Repository status
-
-This is an **early-stage Python repository**. The package scaffold (`src/shard_wiki/`, `tests/`, `pyproject.toml`) exists with only smoke tests — the domain model is not yet implemented. Read `INTENT.md` (aspiration), `SCOPE.md` (current achievement), and `AGENTS.md` (layout and conventions) before designing anything. Close the gap from SCOPE to INTENT via `research/`, `spec/`, and `workplans/`.
-
-## What this project is
-
-`shard-wiki` is a **Git-based Markdown wiki orchestrator and federation layer**, not a wiki engine. It lets multiple heterogeneous wiki-shaped page stores (**shards**) attach to a shared root entity and be presented as a **union of pages**, while preserving each shard's separate storage, provenance, capabilities, and history.
-
-The core job is orchestration across backends — Git repos, repo subdirectories (`wiki/`), Gitea wikis, local folders, Obsidian vaults, WebDAV/Nextcloud directories, Coulomb spaces — never replacing or homogenizing them.
-
-## Core domain model (the concepts code must honor)
-
-These abstractions come from `INTENT.md` and define the architecture. New code should map onto them rather than inventing parallel vocabulary:
-
-- **Shard** — an independently meaningful page store attached to a root entity. Shards have *sovereignty*: their own backend, capabilities, limits, history, and identity model. Not all shards are Git-native.
-- **Root entity / information space** — the joined space that shards attach to. Each information space should have a **Git-addressable coordination layer** (history, patches, review, backup, reconciliation) even when individual shards are not Git-native.
-- **Shard adapter contract** — the versioned interface a backend implements to participate. Adapters are **capability-aware**: the core must model explicitly which operations a shard supports (read, write, diff, merge, lock, version, publish, accept patches) rather than assuming uniformity.
-- **Wiki page model** — a stable, versioned, Markdown-first but backend-neutral representation of pages, paths, links, metadata, revisions.
-- **Projection** — a lazy, cache-like local view of remote/external shard content. Prefer lazy projection over eager copying.
-- **Overlay** — a non-destructive local edit against a remote, read-only, or capability-limited shard, representable as drafts/patches/commits/merge requests *before* destructive application ("overlay before mutation").
-- **Coordination journal** — the Git-backed record of change flows for an information space.
-- **Shard modes** — read-only, write-through, mirrored, projected, cached, canonical.
-
-## Design constraints to enforce in code
-
-These are hard boundaries from `INTENT.md`; treat violations as design bugs:
-
-- **Mechanism over policy.** Provide primitives for federation, sync, overlays, patching, conflict detection, projection, reconciliation. Do *not* hard-code one editorial/sync/conflict/canonical-source policy — keep those configurable.
-- **Union without erasure.** Always preserve provenance: which shard a page came from, its freshness, whether it is cached, whether it has overlays, whether it diverges from an equivalent page elsewhere. Never hide authorship, conflicts, freshness, or backend limitations.
-- **No silent remote mutation.** Do not mutate remote systems without explicit adapter support and user intent.
-- **Graceful degradation.** Limited backends must still be usable as read-only/cache/projection/backup/patch targets.
-- **Not a file-sync daemon.** Synchronization is wiki-page-semantic, not generic file mirroring.
-
-`INTENT.md` has a "Stability Note": changes that redefine what a shard is, Git's role, how root entities are modeled, or whether this is an orchestrator vs. an engine are **architectural changes** and should be rare and deliberate.
-
-## Build, test, run
-
-Python with a `src/` layout, built via hatchling, tested with pytest. Tests run against the source tree directly (`pythonpath = ["src"]` in `pyproject.toml`), so no install/editable step is required to run them.
-
-```bash
-pip install -e ".[dev]"   # one-time: install dev tooling (pytest, pytest-cov, ruff)
-pytest                    # run the full test suite
-pytest tests/test_package.py::test_version_is_exposed  # run a single test
-pytest --cov              # run with coverage
-ruff check                # lint
-ruff format               # format
-```
-
-Note: the system `pytest` is 7.4.x; `minversion` in `pyproject.toml` is pinned to `7.0` to match. Bump it if a newer pytest is installed into the dev environment.
+@SCOPE.md
+@.claude/rules/repo-identity.md
+@.claude/rules/session-protocol.md
+@.claude/rules/first-session.md
+@.claude/rules/workplan-convention.md
+@.claude/rules/stack-and-commands.md
+@.claude/rules/architecture.md
+@.claude/rules/repo-boundary.md
+@.claude/rules/credential-routing.md
+@.claude/rules/agents.md

SCOPE.md

@@ -17,7 +17,7 @@ Learnings update both SCOPE and INTENT where necessary.
 
 | Layer | State |
 |-------|-------|
-| Code | Foundation slice implemented (SHARD-WP-0007): `provenance` + `policy` leaves, `model` (Identity/Placement/Span/Page/CapabilityProfile), `adapters` (contract + FolderAdapter + conformance suite), `coordination` (event-sourced DecisionLog), `union` (resolution + chorus, overlay-aware), `InformationSpace` orchestrator. Write path added (SHARD-WP-0008): writable adapter, overlay engine (draft→patch→apply-under-drift), edit() unifies write-through + overlay-before-mutation. Native engine implemented (SHARD-WP-0014): `engine` (kernel + typed-extension runtime + per-shard activation [ADR-0001] + capability-profile-from-extensions + EngineShardAdapter + the `ext.struct` built-in) — an engine shard attaches to an InformationSpace as a canonical-mode shard. Git-backed coordination log (SHARD-WP-0009): `DecisionLog` storage factored behind an `EventStore`; `GitEventStore` makes the log git-addressable (each space a ref, append = immutable CAS-guarded commit), a per-space `AppendAuthority` (lease) gives a single-writer total order with re-grantable HA hand-off, cross-process read-your-writes verified, and a verbatim one-time importer (`migrate_space`/JSONL) replays in-memory logs into git; `InformationSpace.git_backed(...)` wires it. 128 tests green, ~97% coverage |
+| Code | Foundation slice implemented (SHARD-WP-0007): `provenance` + `policy` leaves, `model` (Identity/Placement/Span/Page/CapabilityProfile), `adapters` (contract + FolderAdapter + conformance suite), `coordination` (event-sourced DecisionLog), `union` (resolution + chorus, overlay-aware), `InformationSpace` orchestrator. Write path added (SHARD-WP-0008): writable adapter, overlay engine (draft→patch→apply-under-drift), edit() unifies write-through + overlay-before-mutation. Native engine implemented (SHARD-WP-0014): `engine` (kernel + typed-extension runtime + per-shard activation [ADR-0001] + capability-profile-from-extensions + EngineShardAdapter + the `ext.struct` built-in) — an engine shard attaches to an InformationSpace as a canonical-mode shard. Git-backed coordination log (SHARD-WP-0009): `DecisionLog` storage factored behind an `EventStore`; `GitEventStore` makes the log git-addressable (each space a ref, append = immutable CAS-guarded commit), a per-space `AppendAuthority` (lease) gives a single-writer total order with re-grantable HA hand-off, cross-process read-your-writes verified, and a verbatim one-time importer (`migrate_space`/JSONL) replays in-memory logs into git; `InformationSpace.git_backed(...)` wires it. Derived views (SHARD-WP-0010): `views` (wikilink + red-link model, BackLinks, RecentChanges, AllPages/SiteMap) — recomputable, provenance-carrying, presentation-free, exposed via `InformationSpace.backlinks/recent_changes/all_pages/site_map`. Incremental-first derived tier (SHARD-WP-0011): `incremental` (indexed equivalence via MinHash/LSH blocking + verify, change-driven delta maintenance with retraction/propagation, Merkle-style digest + self-healing I-2 consistency-checker, `UnionIndex` routed behind `InformationSpace.all_pages` with rebuild as explicit fallback). Second adapter (SHARD-WP-0012): `GitShardAdapter` — git-IS-store substrate (read=tracked *.md, write=commit, current_rev=per-path sha for drift, adopted git-native history), passes conformance, works across folder+git shards in union/overlay/edit with no core change (capability-as-data proven on a second substrate). 196 tests green, ~97% coverage |
 | Intent | `INTENT.md` established; authorization-in-core amendments drafted |
 | Research | yawex prior art; c2 origins; federation concepts; wikiengines overview (`research/260608-*/`); XWiki/TWiki/Foswiki deep dives (`research/260613-*/`); Xanadu + ZigZag + Roam + Obsidian + Notion + Joplin + Logseq + local-first workspaces (Anytype/AFFiNE/AppFlowy) + Trilium + Wiki.js + Federated Wiki + Wikibase + git-forge wikis + TiddlyWiki + ikiwiki + Quip + MojoMojo + Oddmuse + UseModWiki deep dives & shard-spectrum synthesis (`research/260614-*/`) |
 | Demand | NetKingdom integration asks captured, not yet negotiated |

src/shard_wiki/adapters/__init__.py

@@ -9,10 +9,13 @@ from shard_wiki.adapters.conformance import (
 )
 from shard_wiki.adapters.contract import CONTRACT_VERSION, ShardAdapter
 from shard_wiki.adapters.folder import FolderAdapter
+from shard_wiki.adapters.git import GitShardAdapter, PageRevision
 
 __all__ = [
     "ShardAdapter",
     "FolderAdapter",
+    "GitShardAdapter",
+    "PageRevision",
     "CONTRACT_VERSION",
     "Check",
     "ConformanceReport",

src/shard_wiki/adapters/git.py

@@ -0,0 +1,180 @@
+"""GitShardAdapter — a second substrate: git-as-store (SHARD-WP-0012; TSD §A.3 git-IS-store).
+
+The home case where **git is the store *and* the journal**. Tracked ``*.md`` paths are the page
+keys; the working-tree file is the body; a page's ``source_rev`` is the **commit sha of the last
+commit touching its path** (per-path, so an edit to one page never drifts another). The declared
+profile is *git-IS-store ⟹ substrate=git ∧ history=git-native* — the implication rule the
+capability model enforces (§6.5), validated at registration like any other binding.
+
+This adapter adds **no core changes**: it implements the same :class:`ShardAdapter` contract the
+folder adapter does, proving "write an adapter + declare a verified profile" is the whole cost of a
+new substrate (capability-as-data, I-3). Built on the ``git`` CLI via subprocess — zero new deps.
+"""
+
+from __future__ import annotations
+
+import os
+import subprocess
+from collections.abc import Iterable
+from dataclasses import dataclass
+from pathlib import Path
+
+from shard_wiki.adapters.contract import ShardAdapter
+from shard_wiki.model import (
+    AccessGrant,
+    Addressing,
+    AttachmentMode,
+    CapabilityProfile,
+    ContentOpacity,
+    History,
+    Identity,
+    MergeModel,
+    NativeQuery,
+    NotSupported,
+    OperationalEnvelope,
+    Page,
+    Placement,
+    Substrate,
+    Translation,
+    Verb,
+    WriteGranularity,
+)
+from shard_wiki.provenance import Liveness, ProvenanceEnvelope, Staleness
+
+__all__ = ["GitShardAdapter", "PageRevision"]
+
+
+@dataclass(frozen=True, slots=True)
+class PageRevision:
+    """One adopted git-native revision of a page: the commit sha and its subject line."""
+
+    sha: str
+    message: str
+
+_GIT_IDENTITY = {
+    "GIT_AUTHOR_NAME": "shard-wiki",
+    "GIT_AUTHOR_EMAIL": "shard@shard-wiki",
+    "GIT_COMMITTER_NAME": "shard-wiki",
+    "GIT_COMMITTER_EMAIL": "shard@shard-wiki",
+}
+
+
+class GitShardAdapter(ShardAdapter):
+    """A shard whose store is a git repo: keys are tracked ``*.md`` paths, revs are commit shas."""
+
+    def __init__(self, shard_id: str, repo_path: str | Path, writable: bool = False) -> None:
+        self._shard_id = shard_id
+        self._repo = Path(repo_path)
+        self._writable = writable
+        self._repo.mkdir(parents=True, exist_ok=True)
+        if not (self._repo / ".git").exists():
+            self._git("init", "--quiet")
+
+    @property
+    def shard_id(self) -> str:
+        return self._shard_id
+
+    def profile(self) -> CapabilityProfile:
+        # VERSION is always available — a git-IS-store has git-native history to adopt (§A.5),
+        # read-only or not. WRITE (= commit, PER_PAGE) is added only in writable mode.
+        verbs = {Verb.READ, Verb.VERSION}
+        granularity = WriteGranularity.NONE
+        if self._writable:
+            verbs |= {Verb.WRITE}
+            granularity = WriteGranularity.PER_PAGE
+        return CapabilityProfile(
+            substrate=Substrate.GIT,
+            attachment_mode=AttachmentMode.GIT_IS_STORE,
+            write_granularity=granularity,
+            content_opacity=ContentOpacity.TRANSPARENT,
+            operational_envelope=OperationalEnvelope.LOCAL_UNBOUNDED,
+            access_grant=AccessGrant.OPEN,
+            liveness=Liveness.STATIC,
+            history=History.GIT_NATIVE,  # git-is-store ⟹ git-native (§6.5)
+            merge_model=MergeModel.GIT_TEXT,
+            addressing=Addressing.PATH,
+            native_query=NativeQuery.NONE,
+            translation=Translation.NATIVE,
+            supported_verbs=frozenset(verbs),
+        ).validate()
+
+    def write(self, key: str, body: str) -> Page:
+        """Write = **commit**: stage the file and commit it (skip a no-op so no empty commit),
+        returning the page at the new sha. Drift detection rides on ``current_rev`` = that sha."""
+        if not self._writable:
+            raise NotSupported(f"{type(self).__name__} is read-only")
+        rel = f"{key}.md"
+        path = self._path_for(key)
+        path.parent.mkdir(parents=True, exist_ok=True)
+        path.write_text(body, encoding="utf-8")
+        self._git("add", "--", rel)
+        if self._run("diff", "--cached", "--quiet").returncode != 0:  # staged changes present
+            self._git("commit", "-m", f"write {rel}", env=_GIT_IDENTITY)
+        return self.read(key)
+
+    def keys(self) -> Iterable[str]:
+        out = self._git("ls-files", "*.md").decode()
+        for line in out.splitlines():
+            yield line[: -len(".md")] if line.endswith(".md") else line
+
+    def read(self, key: str) -> Page:
+        path = self._path_for(key)
+        if not path.is_file():
+            raise KeyError(key)
+        rev = self.current_rev(key)
+        return Page(
+            identity=Identity(self._shard_id, key),
+            body=path.read_text(encoding="utf-8"),
+            envelope=ProvenanceEnvelope(
+                source_shard=self._shard_id,
+                liveness=Liveness.STATIC,
+                staleness=Staleness.FRESH,
+                source_rev=rev,
+                lineage="git-native",
+            ),
+            placements=(Placement(self._shard_id, f"{key}.md"),),
+        )
+
+    def current_rev(self, key: str) -> str | None:
+        """The sha of the last commit touching ``key``'s path (per-path drift token), or None."""
+        rel = f"{key}.md"
+        if not self._path_for(key).is_file():
+            return None
+        sha = self._git("log", "-1", "--format=%H", "--", rel).decode().strip()
+        return sha or None
+
+    def history(self, key: str) -> tuple[PageRevision, ...]:
+        """Adopt git-native history (§A.5): the commit list for ``key``'s path, newest-first.
+
+        VERSION-gated; raises ``KeyError`` for an unknown page. Each revision is a commit sha +
+        subject — the native log surfaced through the contract, not re-implemented.
+        """
+        if not self.profile().supports(Verb.VERSION):
+            raise NotSupported(f"{type(self).__name__} does not support version")
+        if not self._path_for(key).is_file():
+            raise KeyError(key)
+        out = self._git("log", "--format=%H%x00%s", "--", f"{key}.md").decode()
+        revisions = []
+        for line in out.splitlines():
+            sha, _, message = line.partition("\x00")
+            revisions.append(PageRevision(sha=sha, message=message))
+        return tuple(revisions)
+
+    # -- git plumbing --------------------------------------------------------
+
+    def _path_for(self, key: str) -> Path:
+        return self._repo / f"{key}.md"
+
+    def _git(self, *args: str, stdin: bytes | None = None, env: dict | None = None) -> bytes:
+        return self._run(*args, stdin=stdin, env=env, check=True).stdout
+
+    def _run(
+        self, *args: str, stdin: bytes | None = None, env: dict | None = None, check: bool = False
+    ) -> subprocess.CompletedProcess:
+        return subprocess.run(
+            ["git", "-C", str(self._repo), *args],
+            input=stdin,
+            capture_output=True,
+            env={**os.environ, **(env or {})},
+            check=check,
+        )

src/shard_wiki/incremental/__init__.py

@@ -0,0 +1,46 @@
+"""incremental/ — the incremental-first derived tier (CoreArchitectureBlueprint §8.7).
+
+Equivalence is **indexed** (blocking/LSH + verify), not pairwise O(N²); maintenance is
+**change-driven** (delta with retraction + propagation, review B-4), keeping the derived tier equal
+to a from-scratch rebuild — which becomes a bounded fallback, not the operational path. A
+Merkle-style **digest** plus a background **consistency-checker** make ``derived = f(canonical)``
+verified rather than asserted (I-2), self-healing on detected drift.
+
+In-memory only for this slice (no persisted index store); per-partition structure is honoured but
+multi-tenant deployment is later. Per the dependency rule this imports down (model/provenance) and
+is wired by the orchestrator.
+"""
+
+from shard_wiki.incremental.equivalence import (
+    EquivalenceEdge,
+    EquivalenceIndex,
+    normalized_title,
+)
+from shard_wiki.incremental.minhash import (
+    MinHasher,
+    band_keys,
+    jaccard,
+    shingles,
+)
+from shard_wiki.incremental.union_index import UnionIndex
+from shard_wiki.incremental.verification import (
+    ConsistencyChecker,
+    ConsistencyReport,
+    derived_digest,
+    region_digest,
+)
+
+__all__ = [
+    "shingles",
+    "MinHasher",
+    "band_keys",
+    "jaccard",
+    "EquivalenceEdge",
+    "EquivalenceIndex",
+    "normalized_title",
+    "derived_digest",
+    "region_digest",
+    "ConsistencyReport",
+    "ConsistencyChecker",
+    "UnionIndex",
+]

src/shard_wiki/incremental/equivalence.py

@@ -0,0 +1,225 @@
+"""Indexed equivalence — blocking + verify, incrementally maintained (SHARD-WP-0011 T1/T2).
+
+Equivalence (two *distinct* identities holding the same page) is detected without pairwise O(N²):
+
+1. **Blocking** generates candidate pairs — pages sharing a normalized-title bucket or an LSH band
+   (MinHash over content shingles).
+2. **Verify** confirms a candidate — exact-body fingerprint match, or shingle Jaccard ≥ threshold —
+   plus **curator bindings** (explicit decision-log edges) which are always equivalence edges.
+
+The index is **incrementally maintained** (T2): ``add`` / ``update`` / ``remove`` re-bucket the
+changed page, **retract** the edges it leaves and **add** the edges it enters; equivalence groups
+are the connected components of the current edge set, so a retraction that disconnects a component
+**splits** a chorus automatically. A full :meth:`build` is just repeated ``add`` — the bounded
+rebuild fallback. The invariant (and the test oracle): incremental state == a from-scratch rebuild.
+"""
+
+from __future__ import annotations
+
+import hashlib
+import re
+from collections.abc import Iterable
+from dataclasses import dataclass
+
+from shard_wiki.incremental.minhash import MinHasher, band_keys, jaccard, shingles
+from shard_wiki.model import Identity, Page
+
+__all__ = ["EquivalenceEdge", "EquivalenceIndex", "normalized_title"]
+
+_NONALNUM_RE = re.compile(r"[^a-z0-9]+")
+
+
+def normalized_title(key: str) -> str:
+    """A blocking bucket key: the last path segment, lowercased, stripped of non-alphanumerics."""
+    leaf = key.rsplit("/", 1)[-1]
+    return _NONALNUM_RE.sub("", leaf.lower())
+
+
+@dataclass(frozen=True, slots=True)
+class EquivalenceEdge:
+    """A verified equivalence between two identities, tagged with why it was accepted."""
+
+    a: Identity
+    b: Identity
+    reason: str  # "fingerprint" | "content" | "curator"
+
+
+@dataclass(frozen=True, slots=True)
+class _Entry:
+    shingle_set: frozenset[str]
+    bands: tuple[tuple[int, tuple[int, ...]], ...]
+    title: str
+    fingerprint: str
+
+
+def _fingerprint(body: str) -> str:
+    return hashlib.blake2b(body.strip().encode("utf-8"), digest_size=16).hexdigest()
+
+
+def _pair(a: Identity, b: Identity) -> frozenset[Identity]:
+    return frozenset((a, b))
+
+
+class EquivalenceIndex:
+    """An incrementally maintained, blocked-and-verified equivalence relation over union pages."""
+
+    def __init__(
+        self,
+        *,
+        num_perm: int = 64,
+        num_bands: int = 32,
+        threshold: float = 0.7,
+        hasher: MinHasher | None = None,
+    ) -> None:
+        self.threshold = threshold
+        self.num_bands = num_bands
+        self._hasher = hasher or MinHasher(num_perm=num_perm)
+        self._entries: dict[Identity, _Entry] = {}
+        self._band_buckets: dict[tuple[int, tuple[int, ...]], set[Identity]] = {}
+        self._title_buckets: dict[str, set[Identity]] = {}
+        self._content_edges: dict[frozenset[Identity], str] = {}
+        self._curator_edges: set[frozenset[Identity]] = set()
+
+    # -- build / maintain ----------------------------------------------------
+
+    def build(
+        self,
+        pages: Iterable[Page],
+        curator_edges: Iterable[tuple[Identity, Identity]] = (),
+    ) -> None:
+        """Rebuild from scratch (the bounded fallback): add every page, then curator edges."""
+        self.__init__(
+            num_bands=self.num_bands, threshold=self.threshold, hasher=self._hasher
+        )
+        for page in pages:
+            self.add(page)
+        for a, b in curator_edges:
+            self.bind(a, b)
+
+    def add(self, page: Page) -> None:
+        """Index a new (or, via :meth:`update`, refreshed) page and add its equivalence edges."""
+        identity = page.identity
+        entry = self._make_entry(page)
+        self._entries[identity] = entry
+        for key in entry.bands:
+            self._band_buckets.setdefault(key, set()).add(identity)
+        self._title_buckets.setdefault(entry.title, set()).add(identity)
+
+        for candidate in self._candidates(identity, entry):
+            reason = self._verify(identity, candidate)
+            if reason is not None:
+                self._content_edges[_pair(identity, candidate)] = reason
+
+    def remove(self, identity: Identity) -> None:
+        """Drop a page: de-bucket it and retract every content edge incident to it."""
+        entry = self._entries.pop(identity, None)
+        if entry is None:
+            return
+        for key in entry.bands:
+            self._discard_bucket(self._band_buckets, key, identity)
+        self._discard_bucket(self._title_buckets, entry.title, identity)
+        for edge in [e for e in self._content_edges if identity in e]:
+            del self._content_edges[edge]
+
+    def update(self, page: Page) -> None:
+        """Apply a change as retract-then-add: stale (bucket-exit) edges go, new edges arrive."""
+        self.remove(page.identity)
+        self.add(page)
+
+    def bind(self, a: Identity, b: Identity) -> None:
+        """Record a curator equivalence (an explicit decision-log binding); always an edge."""
+        if a != b:
+            self._curator_edges.add(_pair(a, b))
+
+    def unbind(self, a: Identity, b: Identity) -> None:
+        self._curator_edges.discard(_pair(a, b))
+
+    def set_curator_edges(self, edges: Iterable[tuple[Identity, Identity]]) -> None:
+        """Replace all curator edges at once (re-syncing from the decision-log fold)."""
+        self._curator_edges = {_pair(a, b) for a, b in edges if a != b}
+
+    # -- queries -------------------------------------------------------------
+
+    def identities(self) -> frozenset[Identity]:
+        """All identities currently present in the index."""
+        return frozenset(self._entries)
+
+    def fingerprint(self, identity: Identity) -> str | None:
+        """The content fingerprint indexed for ``identity`` (None if absent) — a digest leaf."""
+        entry = self._entries.get(identity)
+        return entry.fingerprint if entry is not None else None
+
+    def edges(self) -> frozenset[frozenset[Identity]]:
+        """All equivalence edges (content + curator) among currently present identities."""
+        present = self._entries.keys()
+        curator = {e for e in self._curator_edges if e <= present}
+        return frozenset(set(self._content_edges) | curator)
+
+    def groups(self) -> tuple[frozenset[Identity], ...]:
+        """Equivalence groups: connected components of size ≥ 2 (union-find over the edges)."""
+        parent: dict[Identity, Identity] = {}
+
+        def find(x: Identity) -> Identity:
+            parent.setdefault(x, x)
+            root = x
+            while parent[root] != root:
+                root = parent[root]
+            while parent[x] != root:
+                parent[x], x = root, parent[x]
+            return root
+
+        for edge in self.edges():
+            a, b = tuple(edge)
+            ra, rb = find(a), find(b)
+            if ra != rb:
+                parent[ra] = rb
+
+        comps: dict[Identity, set[Identity]] = {}
+        for node in parent:
+            comps.setdefault(find(node), set()).add(node)
+        return tuple(
+            frozenset(members) for members in comps.values() if len(members) > 1
+        )
+
+    def equivalent_to(self, identity: Identity) -> frozenset[Identity]:
+        """The equivalence group containing ``identity`` (including itself), else just itself."""
+        for group in self.groups():
+            if identity in group:
+                return group
+        return frozenset({identity})
+
+    # -- internals -----------------------------------------------------------
+
+    def _make_entry(self, page: Page) -> _Entry:
+        shingle_set = shingles(page.body)
+        signature = self._hasher.signature(shingle_set)
+        return _Entry(
+            shingle_set=shingle_set,
+            bands=band_keys(signature, self.num_bands),
+            title=normalized_title(page.identity.key),
+            fingerprint=_fingerprint(page.body),
+        )
+
+    def _candidates(self, identity: Identity, entry: _Entry) -> set[Identity]:
+        candidates: set[Identity] = set()
+        for key in entry.bands:
+            candidates |= self._band_buckets.get(key, set())
+        candidates |= self._title_buckets.get(entry.title, set())
+        candidates.discard(identity)
+        return candidates
+
+    def _verify(self, a: Identity, b: Identity) -> str | None:
+        ea, eb = self._entries[a], self._entries[b]
+        if ea.fingerprint == eb.fingerprint:
+            return "fingerprint"
+        if jaccard(ea.shingle_set, eb.shingle_set) >= self.threshold:
+            return "content"
+        return None
+
+    @staticmethod
+    def _discard_bucket(buckets: dict, key, identity: Identity) -> None:
+        bucket = buckets.get(key)
+        if bucket is not None:
+            bucket.discard(identity)
+            if not bucket:
+                del buckets[key]

src/shard_wiki/incremental/minhash.py

@@ -0,0 +1,71 @@
+"""MinHash + LSH banding primitives for content-similarity blocking (SHARD-WP-0011 T1).
+
+Pure, deterministic functions (fixed hashing, no per-run randomness) so the derived tier and its
+digest are reproducible. Shingle a body into k-grams, MinHash the shingle set into a signature,
+split the signature into LSH bands; two pages sharing a band are *candidates* for equivalence —
+the cheap pre-filter that replaces pairwise O(N²) comparison.
+"""
+
+from __future__ import annotations
+
+import hashlib
+import random
+import re
+from collections.abc import Iterable
+
+__all__ = ["shingles", "MinHasher", "band_keys", "jaccard"]
+
+_WORD_RE = re.compile(r"\w+")
+# Largest Mersenne prime below 2**61 — the modulus for the universal-hash permutations.
+_PRIME = (1 << 61) - 1
+
+
+def shingles(text: str, k: int = 3) -> frozenset[str]:
+    """The set of word k-grams in ``text`` (lowercased). Short texts fall back to their word set."""
+    words = _WORD_RE.findall(text.lower())
+    if len(words) < k:
+        return frozenset(words)
+    return frozenset(" ".join(words[i : i + k]) for i in range(len(words) - k + 1))
+
+
+def _stable_hash(token: str) -> int:
+    return int.from_bytes(hashlib.blake2b(token.encode("utf-8"), digest_size=8).digest(), "big")
+
+
+class MinHasher:
+    """A bank of ``num_perm`` universal hash permutations producing a fixed-length signature."""
+
+    def __init__(self, num_perm: int = 64, seed: int = 1) -> None:
+        self.num_perm = num_perm
+        rng = random.Random(seed)
+        self._coeffs = [
+            (rng.randrange(1, _PRIME), rng.randrange(0, _PRIME)) for _ in range(num_perm)
+        ]
+
+    def signature(self, shingle_set: Iterable[str]) -> tuple[int, ...]:
+        """The MinHash signature of ``shingle_set`` (empty set → all-``_PRIME`` sentinel)."""
+        hashed = [_stable_hash(s) for s in shingle_set]
+        if not hashed:
+            return tuple(_PRIME for _ in self._coeffs)
+        return tuple(min((a * h + b) % _PRIME for h in hashed) for a, b in self._coeffs)
+
+
+def band_keys(
+    signature: tuple[int, ...], num_bands: int
+) -> tuple[tuple[int, tuple[int, ...]], ...]:
+    """Split a signature into ``num_bands`` band keys; two pages sharing one are LSH candidates."""
+    if num_bands <= 0 or len(signature) % num_bands != 0:
+        raise ValueError(f"signature length {len(signature)} not divisible into {num_bands} bands")
+    rows = len(signature) // num_bands
+    return tuple(
+        (b, signature[b * rows : (b + 1) * rows]) for b in range(num_bands)
+    )
+
+
+def jaccard(a: frozenset[str], b: frozenset[str]) -> float:
+    """Jaccard similarity of two shingle sets; two empty sets are defined as identical (1.0)."""
+    if not a and not b:
+        return 1.0
+    if not a or not b:
+        return 0.0
+    return len(a & b) / len(a | b)

src/shard_wiki/incremental/union_index.py

@@ -0,0 +1,91 @@
+"""UnionIndex — the maintained derived tier wired behind resolution + views (SHARD-WP-0011 T4).
+
+Wraps a :class:`UnionGraph` + decision log with an incrementally maintained
+:class:`EquivalenceIndex`. Content equivalence is kept fresh by deltas (``note_change`` /
+``note_removed``); curator bindings are re-synced live from the log fold. A full :meth:`rebuild`
+is the bounded fallback. :meth:`verify` runs the I-2 consistency-checker over the live source.
+
+Consumer-visible results are unchanged — equivalence groups are exposed in the same string form the
+decision-log fold uses, a *superset* that additionally collapses genuine content duplicates — only
+freshness and cost differ (recompute-on-read becomes change-driven).
+"""
+
+from __future__ import annotations
+
+from shard_wiki.coordination import DecisionLog
+from shard_wiki.incremental.equivalence import EquivalenceIndex
+from shard_wiki.incremental.verification import (
+    ConsistencyChecker,
+    ConsistencyReport,
+    derived_digest,
+)
+from shard_wiki.model import Identity, Page
+from shard_wiki.union import UnionGraph
+
+__all__ = ["UnionIndex"]
+
+
+def _identity(token: str) -> Identity:
+    shard, _, key = token.partition(":")
+    return Identity(shard, key)
+
+
+class UnionIndex:
+    """An incrementally maintained equivalence index over a union, with a rebuild fallback."""
+
+    def __init__(self, union: UnionGraph, log: DecisionLog, space: str) -> None:
+        self._union = union
+        self._log = log
+        self._space = space
+        self._eq = EquivalenceIndex()
+        self.rebuild()
+
+    def rebuild(self) -> None:
+        """The bounded fallback: re-derive the whole index from current union pages + bindings."""
+        self._eq.build(self._union.iter_pages())
+        self._sync_curator()
+
+    def note_change(self, page: Page) -> None:
+        """Change-driven update for one added/edited page (the operational path)."""
+        self._eq.update(page)
+
+    def note_removed(self, identity: Identity) -> None:
+        self._eq.remove(identity)
+
+    def _sync_curator(self) -> None:
+        """Re-sync curator equivalence from the live decision-log fold (cheap, always correct)."""
+        groups = self._log.fold(self._space).equivalence_groups
+        edges: list[tuple[Identity, Identity]] = []
+        for group in groups:
+            members = [_identity(m) for m in group]
+            edges.extend((members[0], other) for other in members[1:])
+        self._eq.set_curator_edges(edges)
+
+    def equivalence_groups(self) -> tuple[frozenset[str], ...]:
+        """Equivalence groups in decision-log string form (curator ∪ content), for the views."""
+        self._sync_curator()
+        return tuple(
+            frozenset(str(identity) for identity in group) for group in self._eq.groups()
+        )
+
+    def digest(self) -> str:
+        """The Merkle-style digest of the maintained derived tier (I-2)."""
+        self._sync_curator()
+        return derived_digest(self._eq)
+
+    def verify(self) -> ConsistencyReport:
+        """Check the maintained index against a from-scratch fold of the live source; self-heal."""
+        self._sync_curator()
+        checker = ConsistencyChecker(
+            self._eq,
+            pages=lambda: list(self._union.iter_pages()),
+            curator_edges=self._curator_pairs,
+        )
+        return checker.check_and_repair()
+
+    def _curator_pairs(self) -> list[tuple[Identity, Identity]]:
+        pairs: list[tuple[Identity, Identity]] = []
+        for group in self._log.fold(self._space).equivalence_groups:
+            members = [_identity(m) for m in group]
+            pairs.extend((members[0], other) for other in members[1:])
+        return pairs

src/shard_wiki/incremental/verification.py

@@ -0,0 +1,112 @@
+"""I-2 verification — digest + background consistency-checker (SHARD-WP-0011 T3).
+
+``derived = f(canonical)`` is made *verified*, not asserted. A **Merkle-style digest** summarizes
+the derived tier (each identity's content fingerprint + its incident equivalence edges as a leaf,
+order-independently combined into a root) so two derived states are equal iff their digests match.
+A **consistency-checker** recomputes the authoritative fold from the current source, compares it to
+the maintained index over a (sampled) region, and on mismatch performs a **scoped recompute** of
+just the affected identities — self-healing drift from a missed delta or corrupted state.
+
+The digest is a pure function of index state, so it is "maintained alongside deltas" for free and
+is stable under equivalent event orders (leaves are sorted before combination).
+"""
+
+from __future__ import annotations
+
+import hashlib
+from collections.abc import Callable, Iterable
+from dataclasses import dataclass
+
+from shard_wiki.incremental.equivalence import EquivalenceIndex
+from shard_wiki.model import Identity, Page
+
+__all__ = ["region_digest", "derived_digest", "ConsistencyReport", "ConsistencyChecker"]
+
+CuratorEdges = Iterable[tuple[Identity, Identity]]
+
+
+def _leaf(index: EquivalenceIndex, identity: Identity) -> str:
+    """A digest leaf for one identity: its fingerprint + its incident edges (as sorted peers)."""
+    fingerprint = index.fingerprint(identity) or "∅"
+    peers = sorted(
+        str(other)
+        for edge in index.edges()
+        if identity in edge
+        for other in edge
+        if other != identity
+    )
+    payload = f"{identity}|{fingerprint}|{','.join(peers)}"
+    return hashlib.blake2b(payload.encode("utf-8"), digest_size=16).hexdigest()
+
+
+def region_digest(index: EquivalenceIndex, identities: Iterable[Identity]) -> str:
+    """A Merkle-style root over the given identities' leaves (order-independent)."""
+    leaves = sorted(_leaf(index, identity) for identity in identities)
+    root = hashlib.blake2b(digest_size=16)
+    for leaf in leaves:
+        root.update(leaf.encode("utf-8"))
+    return root.hexdigest()
+
+
+def derived_digest(index: EquivalenceIndex) -> str:
+    """The digest of the whole maintained derived tier."""
+    return region_digest(index, index.identities())
+
+
+@dataclass(frozen=True, slots=True)
+class ConsistencyReport:
+    """Outcome of a consistency check: what was examined, whether it drifted, and if it healed."""
+
+    checked: int
+    drifted: bool
+    repaired: bool
+    healthy: bool
+
+
+class ConsistencyChecker:
+    """Compares the maintained index against an authoritative rebuild and repairs drift in place."""
+
+    def __init__(
+        self,
+        index: EquivalenceIndex,
+        pages: Callable[[], Iterable[Page]],
+        curator_edges: Callable[[], CuratorEdges] = lambda: (),
+    ) -> None:
+        self._index = index
+        self._pages = pages
+        self._curator = curator_edges
+
+    def _authoritative(self) -> EquivalenceIndex:
+        expected = EquivalenceIndex(
+            num_bands=self._index.num_bands, threshold=self._index.threshold
+        )
+        expected.build(list(self._pages()), list(self._curator()))
+        return expected
+
+    def check_and_repair(self, sample: Iterable[Identity] | None = None) -> ConsistencyReport:
+        """Verify the (sampled) region against a from-scratch fold; scoped-recompute on mismatch."""
+        source = {p.identity: p for p in self._pages()}
+        expected = self._authoritative()
+        region = (
+            set(sample)
+            if sample is not None
+            else set(source) | set(self._index.identities())
+        )
+
+        drifted = region_digest(self._index, region) != region_digest(expected, region)
+        if not drifted:
+            return ConsistencyReport(len(region), drifted=False, repaired=False, healthy=True)
+
+        self._repair(region, source)
+        healthy = region_digest(self._index, region) == region_digest(expected, region)
+        return ConsistencyReport(len(region), drifted=True, repaired=True, healthy=healthy)
+
+    def _repair(self, region: set[Identity], source: dict[Identity, Page]) -> None:
+        """Scoped recompute: reconcile each affected identity to the current source."""
+        present = self._index.identities()
+        for identity in region:
+            page = source.get(identity)
+            if page is not None:
+                self._index.update(page) if identity in present else self._index.add(page)
+            elif identity in present:
+                self._index.remove(identity)

src/shard_wiki/space.py

@@ -20,9 +20,20 @@ from shard_wiki.coordination import (
     Overlay,
     OverlayEngine,
 )
+from shard_wiki.incremental import ConsistencyReport, UnionIndex
 from shard_wiki.model import Page
 from shard_wiki.policy import DEFAULT_POLICY, Policy
 from shard_wiki.union import Resolution, UnionGraph
+from shard_wiki.views import (
+    AllPagesEntry,
+    BackLink,
+    ChangeEntry,
+    SiteMapNode,
+    all_pages,
+    build_backlinks,
+    recent_changes,
+    site_map,
+)
 
 __all__ = ["InformationSpace"]
 
@@ -41,6 +52,8 @@ class InformationSpace:
         self.log = DecisionLog(store)
         self.union = UnionGraph(space_id, log=self.log, policy=policy)
         self.overlays = OverlayEngine(space_id, self.log)
+        self._index: UnionIndex | None = None  # maintained derived tier, built lazily
+        self._index_stale = True
 
     @classmethod
     def git_backed(
@@ -57,6 +70,7 @@ class InformationSpace:
         """Attach a shard — only if it passes conformance (verified profile, I-3/§6.6)."""
         assert_conformant(adapter)
         self.union.attach(adapter)
+        self._index_stale = True
 
     def alias(self, name: str, target: str, actor: str | None = None) -> None:
         """Record a coordination-canonical alias (``name`` → ``"shard:key"``) in the log."""
@@ -91,4 +105,44 @@ class InformationSpace:
         write-through-capable target fast-forwards (write-through); a read-only target keeps the
         draft as local truth (I-5: overlay before mutation, always)."""
         overlay = self.overlay(name, body, actor=actor)
-        return self.apply_overlay(overlay.overlay_id)
+        result = self.apply_overlay(overlay.overlay_id)
+        self._index_stale = True  # the applied edit changes the derived tier
+        return result
+
+    # --- maintained derived tier (SHARD-WP-0011): incremental-first, rebuild as fallback ---
+
+    @property
+    def index(self) -> UnionIndex:
+        """The maintained equivalence index (built lazily; rebuilt when the union has changed)."""
+        if self._index is None:
+            self._index = UnionIndex(self.union, self.log, self.space_id)
+        elif self._index_stale:
+            self._index.rebuild()  # bounded fallback after a mutation
+        self._index_stale = False
+        return self._index
+
+    def reindex(self) -> None:
+        """Force a full rebuild of the maintained derived tier (the explicit fallback path)."""
+        self.index.rebuild()
+
+    def verify_index(self) -> ConsistencyReport:
+        """Run the I-2 consistency-checker over the maintained tier; self-heal any drift."""
+        return self.index.verify()
+
+    # --- derived views (SHARD-WP-0010): recomputable, provenance-carrying, presentation-free ---
+
+    def backlinks(self, name: str, *, camelcase: bool = False) -> tuple[BackLink, ...]:
+        """Pages across the union that link to ``name`` (UC-18)."""
+        return build_backlinks(self.union, camelcase=camelcase).to(name)
+
+    def recent_changes(self, *, limit: int | None = None) -> tuple[ChangeEntry, ...]:
+        """The merged newest-first change feed: coordination journal + shard signals (UC-17)."""
+        return recent_changes(self.union, self.log, self.space_id, limit=limit)
+
+    def all_pages(self) -> tuple[AllPagesEntry, ...]:
+        """The union's distinct pages, collapsed via the maintained equivalence index."""
+        return all_pages(self.union, equivalence_groups=self.index.equivalence_groups())
+
+    def site_map(self) -> SiteMapNode:
+        """The union namespace tree built from page placements."""
+        return site_map(self.union)

src/shard_wiki/union/resolver.py

@@ -13,6 +13,7 @@ imported by nothing.
 from __future__ import annotations
 
 import dataclasses
+from collections.abc import Iterator
 from dataclasses import dataclass
 from enum import Enum
 
@@ -68,6 +69,20 @@ class UnionGraph:
     def shard(self, shard_id: str) -> ShardAdapter | None:
         return next((s for s in self._shards if s.shard_id == shard_id), None)
 
+    @property
+    def shards(self) -> tuple[ShardAdapter, ...]:
+        return tuple(self._shards)
+
+    def iter_pages(self) -> Iterator[Page]:
+        """Every page across attached shards, raw (per-shard, not chorus-collapsed). The
+        enumeration substrate for derived views — BackLinks, AllPages, SiteMap (§8.4)."""
+        for shard in self._shards:
+            for key in shard.keys():
+                try:
+                    yield shard.read(key)
+                except KeyError:
+                    continue
+
     def _read_all(self, key: str) -> list[Page]:
         pages: list[Page] = []
         for shard in self._shards:

src/shard_wiki/views/__init__.py

@@ -0,0 +1,33 @@
+"""views/ — derived, recomputable, provenance-carrying read views over the union (§8.4).
+
+All views here are *derived tier*: pure functions of the attached shards plus the coordination-log
+fold, storing nothing canonical (SHARD-WP-0011 makes them incrementally maintainable). Presentation
+stays out of core (L6) — these produce models, never rendered output. Per the dependency rule this
+package imports down (union/model/coordination/provenance) and is imported only by the orchestrator.
+"""
+
+from shard_wiki.views.allpages import AllPagesEntry, SiteMapNode, all_pages, site_map
+from shard_wiki.views.backlinks import BackLink, BackLinksIndex, build_backlinks
+from shard_wiki.views.links import (
+    ResolvedLink,
+    WikiLink,
+    extract_links,
+    resolve_links,
+)
+from shard_wiki.views.recentchanges import ChangeEntry, recent_changes
+
+__all__ = [
+    "WikiLink",
+    "ResolvedLink",
+    "extract_links",
+    "resolve_links",
+    "BackLink",
+    "BackLinksIndex",
+    "build_backlinks",
+    "ChangeEntry",
+    "recent_changes",
+    "AllPagesEntry",
+    "SiteMapNode",
+    "all_pages",
+    "site_map",
+]

src/shard_wiki/views/allpages.py

@@ -0,0 +1,131 @@
+"""AllPages + SiteMap — enumeration views over the union (SHARD-WP-0010 T4).
+
+**AllPages** lists the union's distinct pages, collapsing identities that name the same page: a
+*chorus* (same key across shards) and *equivalence-bound* identities (decision-log bindings) fold
+into one entry, with divergence noted when the members' bodies differ (union without erasure — the
+collapse is acknowledged, never silent). **SiteMap** is the namespace tree built from page
+placements (paths), spanning shards.
+
+Both are derived/recomputable and presentation-free (the tree is a model, not rendered HTML).
+"""
+
+from __future__ import annotations
+
+from dataclasses import dataclass
+
+from shard_wiki.model import Identity, Page
+from shard_wiki.union import UnionGraph
+
+__all__ = ["AllPagesEntry", "SiteMapNode", "all_pages", "site_map"]
+
+
+@dataclass(frozen=True, slots=True)
+class AllPagesEntry:
+    """One union page: its representative ``name``, the ``members`` collapsed into it, and whether
+    those members' bodies ``diverge`` (a chorus with differing content)."""
+
+    name: str
+    members: tuple[Identity, ...]
+    diverges: bool
+
+
+@dataclass(frozen=True, slots=True)
+class SiteMapNode:
+    """A namespace node: its path ``name``, child namespaces, and pages directly under it."""
+
+    name: str
+    children: tuple[SiteMapNode, ...]
+    pages: tuple[Identity, ...]
+
+
+class _UnionFind:
+    def __init__(self) -> None:
+        self._parent: dict[str, str] = {}
+
+    def add(self, x: str) -> None:
+        self._parent.setdefault(x, x)
+
+    def find(self, x: str) -> str:
+        self.add(x)
+        root = x
+        while self._parent[root] != root:
+            root = self._parent[root]
+        while self._parent[x] != root:
+            self._parent[x], x = root, self._parent[x]
+        return root
+
+    def union(self, a: str, b: str) -> None:
+        self.add(a)
+        self.add(b)
+        ra, rb = self.find(a), self.find(b)
+        if ra != rb:
+            self._parent[max(ra, rb)] = min(ra, rb)
+
+
+def all_pages(
+    union: UnionGraph,
+    equivalence_groups: tuple[frozenset[str], ...] | None = None,
+) -> tuple[AllPagesEntry, ...]:
+    """Enumerate the union's distinct pages, collapsing chorus + equivalence-bound members.
+
+    ``equivalence_groups`` (string identities, decision-log form) overrides the source of
+    equivalence — the orchestrator passes the maintained index's groups (SHARD-WP-0011 T4); the
+    default falls back to the decision-log fold, so direct callers are unaffected.
+    """
+    pages: dict[str, Page] = {}
+    by_key: dict[str, list[str]] = {}
+    for page in union.iter_pages():
+        ident = str(page.identity)
+        pages[ident] = page
+        by_key.setdefault(page.identity.key, []).append(ident)
+
+    uf = _UnionFind()
+    for ident in pages:
+        uf.add(ident)
+    for idents in by_key.values():  # same key across shards → chorus
+        for other in idents[1:]:
+            uf.union(idents[0], other)
+    if equivalence_groups is None:
+        equivalence_groups = union.log.fold(union.space).equivalence_groups
+    for group in equivalence_groups:  # curator bindings (+ maintained content edges)
+        present = [m for m in group if m in pages]
+        for other in present[1:]:
+            uf.union(present[0], other)
+
+    groups: dict[str, list[str]] = {}
+    for ident in pages:
+        groups.setdefault(uf.find(ident), []).append(ident)
+
+    entries: list[AllPagesEntry] = []
+    for members in groups.values():
+        member_pages = [pages[m] for m in members]
+        identities = tuple(p.identity for p in member_pages)
+        name = min(p.identity.key for p in member_pages)
+        diverges = len({p.body for p in member_pages}) > 1
+        entries.append(AllPagesEntry(name=name, members=identities, diverges=diverges))
+    return tuple(sorted(entries, key=lambda e: e.name))
+
+
+def _segments(page: Page) -> list[str]:
+    path = page.placements[0].path if page.placements else page.identity.key
+    if path.endswith(".md"):
+        path = path[:-3]
+    return [seg for seg in path.split("/") if seg]
+
+
+def site_map(union: UnionGraph) -> SiteMapNode:
+    """The union namespace tree from page placements (directories nest; pages sit at their dir)."""
+    root: dict = {"children": {}, "pages": []}
+    for page in union.iter_pages():
+        segments = _segments(page)
+        node = root
+        for seg in segments[:-1]:  # directory segments build the nesting
+            node = node["children"].setdefault(seg, {"children": {}, "pages": []})
+        node["pages"].append(page.identity)
+    return _freeze("", root)
+
+
+def _freeze(name: str, node: dict) -> SiteMapNode:
+    children = tuple(_freeze(k, v) for k, v in sorted(node["children"].items()))
+    pages = tuple(sorted(node["pages"], key=str))
+    return SiteMapNode(name=name, children=children, pages=pages)

src/shard_wiki/views/backlinks.py

@@ -0,0 +1,65 @@
+"""BackLinks — the strongest core derived view (SHARD-WP-0010 T2; UC-18).
+
+For any page name, the set of pages that link to it. Built by extracting wikilinks (T1) from every
+page across the attached shards and resolving each through the union: only **resolved** links
+create a backlink (a red-link points at nothing, so it contributes none). Entries carry their
+**source provenance** (the linking page's identity / shard). Keying by the resolved *name* means a
+chorus target aggregates the backlinks of all its members into one bucket (union without erasure).
+
+Derived/recomputable — stores nothing canonical; SHARD-WP-0011 maintains it incrementally.
+"""
+
+from __future__ import annotations
+
+from collections.abc import Mapping
+from dataclasses import dataclass
+
+from shard_wiki.model import Identity
+from shard_wiki.union import UnionGraph
+from shard_wiki.views.links import resolve_links
+
+__all__ = ["BackLink", "BackLinksIndex", "build_backlinks"]
+
+
+@dataclass(frozen=True, slots=True)
+class BackLink:
+    """One inbound link: ``source`` (the linking page) references ``target_name``."""
+
+    source: Identity
+    target_name: str
+
+    @property
+    def source_shard(self) -> str:
+        return self.source.shard
+
+
+class BackLinksIndex:
+    """An immutable name → inbound-links index over the union link graph."""
+
+    def __init__(self, edges: Mapping[str, tuple[BackLink, ...]]) -> None:
+        self._edges = dict(edges)
+
+    def to(self, name: str) -> tuple[BackLink, ...]:
+        """The backlinks pointing at ``name`` (empty if none)."""
+        return self._edges.get(name, ())
+
+    def sources(self, name: str) -> frozenset[Identity]:
+        """Just the identities linking to ``name`` — convenient for set assertions."""
+        return frozenset(bl.source for bl in self.to(name))
+
+    def names(self) -> frozenset[str]:
+        return frozenset(self._edges)
+
+
+def build_backlinks(union: UnionGraph, *, camelcase: bool = False) -> BackLinksIndex:
+    """Scan every union page's links and index the resolved ones by target name."""
+    edges: dict[str, set[BackLink]] = {}
+    for page in union.iter_pages():
+        for resolved in resolve_links(union, page.body, camelcase=camelcase):
+            if resolved.is_red_link:
+                continue  # red-links don't create backlinks
+            backlink = BackLink(source=page.identity, target_name=resolved.link.target)
+            edges.setdefault(resolved.link.target, set()).add(backlink)
+    return BackLinksIndex(
+        {name: tuple(sorted(links, key=lambda bl: str(bl.source))) for name, links in edges.items()}
+    )

src/shard_wiki/views/links.py

@@ -0,0 +1,91 @@
+"""Wikilink + red-link model (SHARD-WP-0010 T1; FederationRequirements ADR-06).
+
+A CommonMark *wikilink extension*: ``[[Target]]`` and ``[[Target|label]]`` are extracted from a
+page body and each target is resolved through the union (ADR-01). A target that resolves is a
+**link**; one that does not is a **red-link** — a createable hole (UC-23), never a dropped
+reference (union without erasure). CamelCase auto-linking (``WikiWord``) is **off by default** and
+opt-in per space, since bare CamelCase is noisy and policy-laden.
+
+The link *model and resolution* are core; turning a :class:`ResolvedLink` into an ``<a>`` (or a
+red anchor) is L6 presentation and lives outside this package. Link spans are byte/char offsets in
+the body so a later layer can address them precisely.
+"""
+
+from __future__ import annotations
+
+import re
+from dataclasses import dataclass
+
+from shard_wiki.union import Resolution, UnionGraph
+
+__all__ = ["WikiLink", "ResolvedLink", "extract_links", "resolve_links"]
+
+_WIKILINK_RE = re.compile(r"\[\[\s*([^\]|]+?)\s*(?:\|\s*([^\]]+?)\s*)?\]\]")
+# A WikiWord: ≥2 capitalized alphanumeric segments run together (e.g. FrontPage, WikiWord).
+_CAMELCASE_RE = re.compile(r"\b([A-Z][a-z0-9]+(?:[A-Z][a-z0-9]+)+)\b")
+_FENCED_RE = re.compile(r"```.*?```", re.DOTALL)
+_INLINE_CODE_RE = re.compile(r"`[^`\n]*`")
+
+
+@dataclass(frozen=True, slots=True)
+class WikiLink:
+    """One extracted reference. ``target`` is the resolve key; ``label`` is the display text (or
+    None to use the target); ``span`` is the ``[start, end)`` offset of the whole token in the body;
+    ``auto`` marks a CamelCase auto-link (vs an explicit ``[[...]]``)."""
+
+    target: str
+    label: str | None
+    span: tuple[int, int]
+    auto: bool = False
+
+    @property
+    def text(self) -> str:
+        return self.label or self.target
+
+
+@dataclass(frozen=True, slots=True)
+class ResolvedLink:
+    """A :class:`WikiLink` paired with its union :class:`Resolution` (the link's truth status)."""
+
+    link: WikiLink
+    resolution: Resolution
+
+    @property
+    def is_red_link(self) -> bool:
+        return self.resolution.is_red_link
+
+
+def _mask(body: str, pattern: re.Pattern[str]) -> str:
+    """Blank out ``pattern`` matches with equal-length spaces so later scans skip them while every
+    surviving match keeps its true offset."""
+    return pattern.sub(lambda m: " " * len(m.group(0)), body)
+
+
+def extract_links(body: str, *, camelcase: bool = False) -> tuple[WikiLink, ...]:
+    """Extract wikilinks from ``body`` in document order, skipping fenced/inline code.
+
+    With ``camelcase=True`` (per-space opt-in), bare ``WikiWord`` tokens outside code and outside
+    existing ``[[...]]`` also become links.
+    """
+    scan = _mask(_mask(body, _FENCED_RE), _INLINE_CODE_RE)
+    links: list[WikiLink] = []
+    for m in _WIKILINK_RE.finditer(scan):
+        links.append(WikiLink(target=m.group(1).strip(), label=m.group(2), span=m.span()))
+
+    if camelcase:
+        # Mask explicit-link spans too, so a CamelCase target inside [[...]] isn't double-counted.
+        cc_scan = _mask(scan, _WIKILINK_RE)
+        for m in _CAMELCASE_RE.finditer(cc_scan):
+            links.append(WikiLink(target=m.group(1), label=None, span=m.span(), auto=True))
+
+    return tuple(sorted(links, key=lambda link: link.span[0]))
+
+
+def resolve_links(
+    union: UnionGraph, body: str, *, camelcase: bool = False
+) -> tuple[ResolvedLink, ...]:
+    """Extract and resolve every link in ``body`` against ``union`` (link vs red-link, ADR-01)."""
+    return tuple(
+        ResolvedLink(link, union.resolve(link.target))
+        for link in extract_links(body, camelcase=camelcase)
+    )

src/shard_wiki/views/recentchanges.py

@@ -0,0 +1,108 @@
+"""RecentChanges — a merged change feed over the union (SHARD-WP-0010 T3; UC-17).
+
+Two streams, one ordered feed (newest-first):
+
+* the **coordination journal** — overlay/alias/fork/merge/binding decisions from the decision log,
+  each carrying its actor and the decision payload; and
+* **shard change signals** — a page's current revision (folder mtime / ``source_rev``), i.e. the
+  backend's own "this changed" evidence.
+
+Every entry carries provenance: which shard the edit came from, or that it was a coordination
+decision (and by whom). Derived/recomputable — `notify`-driven streaming is a later binding.
+"""
+
+from __future__ import annotations
+
+from collections.abc import Mapping
+from dataclasses import dataclass, field
+from datetime import datetime
+
+from shard_wiki.coordination import DecisionLog, EventType
+from shard_wiki.union import UnionGraph
+
+__all__ = ["ChangeEntry", "recent_changes"]
+
+_COORDINATION = "coordination"
+
+# How each journal event names the thing it touched + a human kind label.
+_EVENT_KIND = {
+    EventType.ALIAS_SET: "alias",
+    EventType.OVERLAY_CREATED: "overlay",
+    EventType.MERGE_DECIDED: "merge",
+    EventType.PAGE_FORKED: "fork",
+    EventType.BINDING_MADE: "binding",
+}
+
+
+@dataclass(frozen=True, slots=True)
+class ChangeEntry:
+    """One change in the feed. ``source`` is the shard id (a shard edit) or ``"coordination"``."""
+
+    when: datetime
+    kind: str
+    ref: str
+    source: str
+    actor: str | None = None
+    detail: Mapping[str, object] = field(default_factory=dict)
+
+
+def _event_ref(event_type: EventType, payload: Mapping[str, object]) -> str:
+    if event_type is EventType.ALIAS_SET:
+        return str(payload.get("alias", ""))
+    if event_type is EventType.OVERLAY_CREATED:
+        return f"{payload.get('target_shard')}:{payload.get('target_key')}"
+    if event_type is EventType.PAGE_FORKED:
+        return f"{payload.get('source')}→{payload.get('fork')}"
+    if event_type is EventType.BINDING_MADE:
+        return ", ".join(str(m) for m in payload.get("members", ()))
+    return str(payload.get("overlay_id", ""))  # MERGE_DECIDED
+
+
+def recent_changes(
+    union: UnionGraph,
+    log: DecisionLog,
+    space: str,
+    *,
+    limit: int | None = None,
+) -> tuple[ChangeEntry, ...]:
+    """Merge the coordination journal and shard change signals into one newest-first feed."""
+    entries: list[ChangeEntry] = []
+
+    for event in log.events(space):
+        entries.append(
+            ChangeEntry(
+                when=event.timestamp,
+                kind=_EVENT_KIND.get(event.type, event.type.value),
+                ref=_event_ref(event.type, event.payload),
+                source=_COORDINATION,
+                actor=event.actor,
+                detail=dict(event.payload),
+            )
+        )
+
+    for page in union.iter_pages():
+        rev = page.envelope.source_rev
+        when = _parse_rev(rev)
+        if when is None:
+            continue  # shard offers no change signal for this page — skip gracefully
+        entries.append(
+            ChangeEntry(
+                when=when,
+                kind="edit",
+                ref=str(page.identity),
+                source=page.identity.shard,
+                detail={"source_rev": rev},
+            )
+        )
+
+    entries.sort(key=lambda e: e.when, reverse=True)
+    return tuple(entries if limit is None else entries[:limit])
+
+
+def _parse_rev(rev: str | None) -> datetime | None:
+    if rev is None:
+        return None
+    try:
+        return datetime.fromisoformat(rev)
+    except ValueError:
+        return None  # non-temporal revision token (e.g. a content hash) — no feed timestamp

tests/test_git_adapter.py

@@ -0,0 +1,131 @@
+"""Tests for the GitShardAdapter read path + profile (SHARD-WP-0012 T1)."""
+
+import subprocess
+
+import pytest
+
+from shard_wiki.adapters import GitShardAdapter, run_conformance
+from shard_wiki.model import (
+    AttachmentMode,
+    History,
+    NotSupported,
+    ProfileError,
+    Substrate,
+    Verb,
+)
+
+
+def _git(repo, *args):
+    subprocess.run(
+        ["git", "-C", str(repo), *args],
+        check=True,
+        capture_output=True,
+        env={"GIT_AUTHOR_NAME": "t", "GIT_AUTHOR_EMAIL": "t@t",
+             "GIT_COMMITTER_NAME": "t", "GIT_COMMITTER_EMAIL": "t@t",
+             "PATH": __import__("os").environ.get("PATH", "")},
+    )
+
+
+def _repo(tmp_path, files, name="repo"):
+    repo = tmp_path / name
+    repo.mkdir()
+    _git(repo, "init", "--quiet")
+    for rel, text in files.items():
+        p = repo / rel
+        p.parent.mkdir(parents=True, exist_ok=True)
+        p.write_text(text, encoding="utf-8")
+        _git(repo, "add", rel)
+    _git(repo, "commit", "-m", "seed")
+    return repo
+
+
+def test_keys_are_tracked_md_paths(tmp_path):
+    repo = _repo(tmp_path, {"Home.md": "h", "docs/Guide.md": "g", "ignore.txt": "x"})
+    adapter = GitShardAdapter("git", repo)
+    assert set(adapter.keys()) == {"Home", "docs/Guide"}  # only tracked *.md
+
+
+def test_read_returns_page_with_commit_sha_rev(tmp_path):
+    repo = _repo(tmp_path, {"Home.md": "welcome"})
+    adapter = GitShardAdapter("git", repo)
+    page = adapter.read("Home")
+    assert page.identity.shard == "git"
+    assert page.body == "welcome"
+    head = subprocess.run(
+        ["git", "-C", str(repo), "rev-parse", "HEAD"], capture_output=True, text=True, check=True
+    ).stdout.strip()
+    assert page.envelope.source_rev == head  # source_rev is the commit sha
+    assert page.envelope.lineage == "git-native"
+
+
+def test_read_missing_key_raises(tmp_path):
+    adapter = GitShardAdapter("git", _repo(tmp_path, {"Home.md": "h"}))
+    with pytest.raises(KeyError):
+        adapter.read("Nope")
+
+
+def test_profile_validates_implication_rules(tmp_path):
+    profile = GitShardAdapter("git", _repo(tmp_path, {"Home.md": "h"})).profile()
+    assert profile.substrate is Substrate.GIT
+    assert profile.attachment_mode is AttachmentMode.GIT_IS_STORE
+    assert profile.history is History.GIT_NATIVE  # git-is-store ⟹ git-native
+    profile.validate()  # raises if the implication rule were violated
+
+
+def test_profile_is_read_only_in_t1(tmp_path):
+    profile = GitShardAdapter("git", _repo(tmp_path, {"Home.md": "h"})).profile()
+    assert profile.supports(Verb.READ)
+    assert not profile.supports(Verb.WRITE)
+
+
+def test_conformance_read_path_passes(tmp_path):
+    adapter = GitShardAdapter("git", _repo(tmp_path, {"Home.md": "h", "Other.md": "o"}))
+    report = run_conformance(adapter)
+    assert report.ok, report.diff()
+
+
+def test_unclaimed_write_raises_not_supported(tmp_path):
+    adapter = GitShardAdapter("git", _repo(tmp_path, {"Home.md": "h"}))
+    with pytest.raises(NotSupported):
+        adapter.write("Home", "new")  # read-only: honest absence
+
+
+def test_empty_repo_has_no_keys(tmp_path):
+    repo = tmp_path / "empty"
+    repo.mkdir()
+    _git(repo, "init", "--quiet")
+    adapter = GitShardAdapter("git", repo)
+    assert list(adapter.keys()) == []
+
+
+def test_bad_profile_combo_is_rejected():
+    # Sanity: the implication rule that backs the git profile actually bites when violated.
+    from shard_wiki.model import (
+        AccessGrant,
+        Addressing,
+        CapabilityProfile,
+        ContentOpacity,
+        MergeModel,
+        NativeQuery,
+        OperationalEnvelope,
+        Translation,
+        WriteGranularity,
+    )
+    from shard_wiki.provenance import Liveness
+
+    with pytest.raises(ProfileError):
+        CapabilityProfile(
+            substrate=Substrate.FILES,  # not git, but claims git-is-store
+            attachment_mode=AttachmentMode.GIT_IS_STORE,
+            write_granularity=WriteGranularity.NONE,
+            content_opacity=ContentOpacity.TRANSPARENT,
+            operational_envelope=OperationalEnvelope.LOCAL_UNBOUNDED,
+            access_grant=AccessGrant.OPEN,
+            liveness=Liveness.STATIC,
+            history=History.NONE,
+            merge_model=MergeModel.NONE,
+            addressing=Addressing.PATH,
+            native_query=NativeQuery.NONE,
+            translation=Translation.NATIVE,
+            supported_verbs=frozenset({Verb.READ}),
+        ).validate()

tests/test_git_adapter_integration.py

@@ -0,0 +1,116 @@
+"""GitShardAdapter history adopt + cross-substrate integration (SHARD-WP-0012 T3)."""
+
+import os
+import subprocess
+
+import pytest
+
+from shard_wiki.adapters import FolderAdapter, GitShardAdapter
+from shard_wiki.coordination import ApplyStatus
+from shard_wiki.space import InformationSpace
+
+_ENV = {
+    "GIT_AUTHOR_NAME": "t", "GIT_AUTHOR_EMAIL": "t@t",
+    "GIT_COMMITTER_NAME": "t", "GIT_COMMITTER_EMAIL": "t@t",
+    "PATH": os.environ.get("PATH", ""),
+}
+
+
+def _git(repo, *args):
+    return subprocess.run(
+        ["git", "-C", str(repo), *args], check=True, capture_output=True, text=True, env=_ENV
+    ).stdout.strip()
+
+
+def _git_repo(tmp_path, files, name="git"):
+    repo = tmp_path / name
+    repo.mkdir()
+    _git(repo, "init", "--quiet")
+    for rel, text in files.items():
+        (repo / rel).parent.mkdir(parents=True, exist_ok=True)
+        (repo / rel).write_text(text, encoding="utf-8")
+        _git(repo, "add", rel)
+    _git(repo, "commit", "-m", "seed")
+    return repo
+
+
+def _folder(tmp_path, name, files, writable=False):
+    root = tmp_path / name
+    for rel, text in files.items():
+        p = root / rel
+        p.parent.mkdir(parents=True, exist_ok=True)
+        p.write_text(text, encoding="utf-8")
+    return FolderAdapter(name, root, writable=writable)
+
+
+# -- history adopt -------------------------------------------------------------
+
+
+def test_history_lists_commits_newest_first(tmp_path):
+    repo = _git_repo(tmp_path, {"Home.md": "v1"})
+    adapter = GitShardAdapter("git", repo, writable=True)
+    adapter.write("Home", "v2")
+    history = adapter.history("Home")
+    assert len(history) == 2
+    assert history[0].message == "write Home.md"  # newest first
+    assert history[-1].message == "seed"
+    assert all(rev.sha for rev in history)
+
+
+def test_history_unknown_key_raises(tmp_path):
+    adapter = GitShardAdapter("git", _git_repo(tmp_path, {"Home.md": "h"}))
+    with pytest.raises(KeyError):
+        adapter.history("Nope")
+
+
+# -- cross-substrate integration ----------------------------------------------
+
+
+def test_resolve_across_git_and_folder(tmp_path):
+    space = InformationSpace("space")
+    space.attach(GitShardAdapter("git", _git_repo(tmp_path, {"Home.md": "git home"})))
+    space.attach(_folder(tmp_path, "notes", {"Daily.md": "folder daily"}))
+    assert space.read("Home").body == "git home"  # resolved from the git shard
+    assert space.read("Daily").body == "folder daily"  # resolved from the folder shard
+
+
+def test_chorus_spans_substrates_with_divergence(tmp_path):
+    space = InformationSpace("space")
+    space.attach(GitShardAdapter("git", _git_repo(tmp_path, {"Shared.md": "from git"})))
+    space.attach(_folder(tmp_path, "notes", {"Shared.md": "from folder"}))
+    res = space.resolve("Shared")
+    assert {p.body for p in res.pages} == {"from git", "from folder"}  # chorus across substrates
+    git_page = next(p for p in res.pages if p.identity.shard == "git")
+    assert git_page.envelope.divergence  # divergence recorded, not erased
+
+
+def test_edit_through_git_shard_commits(tmp_path):
+    repo = _git_repo(tmp_path, {"Home.md": "original"})
+    space = InformationSpace("space")
+    space.attach(GitShardAdapter("git", repo, writable=True))
+    result = space.edit("Home", "edited via overlay")
+    assert result.status is ApplyStatus.APPLIED  # write-through fast-forward on a git shard
+    assert space.read("Home").body == "edited via overlay"
+    assert int(_git(repo, "rev-list", "--count", "HEAD")) == 2  # the edit became a commit
+
+
+def test_apply_under_drift_refuses_on_external_commit(tmp_path):
+    repo = _git_repo(tmp_path, {"Home.md": "original"})
+    space = InformationSpace("space")
+    space.attach(GitShardAdapter("git", repo, writable=True))
+    overlay = space.overlay("Home", "my draft")  # base_rev = current git sha
+    # Another writer commits to the same path → the sha moves underneath the draft.
+    (repo / "Home.md").write_text("someone else", encoding="utf-8")
+    _git(repo, "add", "Home.md")
+    _git(repo, "commit", "-m", "external")
+    result = space.apply_overlay(overlay.overlay_id)
+    assert result.status is ApplyStatus.REFUSED_DRIFT  # never clobber (sha drift detected)
+    # The shard itself is untouched — the external commit stands; the draft remains a draft.
+    assert space.union.shard("git").read("Home").body == "someone else"
+
+
+def test_overlay_on_read_only_git_shard_kept_as_draft(tmp_path):
+    space = InformationSpace("space")
+    space.attach(GitShardAdapter("git", _git_repo(tmp_path, {"Home.md": "ro"}), writable=False))
+    result = space.edit("Home", "wanted change")
+    assert result.status is ApplyStatus.KEPT_DRAFT  # read-only target → overlay retained

tests/test_git_adapter_write.py

@@ -0,0 +1,89 @@
+"""Tests for GitShardAdapter write=commit + current_rev drift (SHARD-WP-0012 T2)."""
+
+import os
+import subprocess
+
+from shard_wiki.adapters import GitShardAdapter, run_conformance
+from shard_wiki.model import Verb
+
+_ENV = {
+    "GIT_AUTHOR_NAME": "t", "GIT_AUTHOR_EMAIL": "t@t",
+    "GIT_COMMITTER_NAME": "t", "GIT_COMMITTER_EMAIL": "t@t",
+    "PATH": os.environ.get("PATH", ""),
+}
+
+
+def _git(repo, *args, capture=False):
+    return subprocess.run(
+        ["git", "-C", str(repo), *args], check=True, capture_output=True, text=True, env=_ENV
+    ).stdout.strip()
+
+
+def _repo(tmp_path, files):
+    repo = tmp_path / "repo"
+    repo.mkdir()
+    _git(repo, "init", "--quiet")
+    for rel, text in files.items():
+        (repo / rel).write_text(text, encoding="utf-8")
+        _git(repo, "add", rel)
+    _git(repo, "commit", "-m", "seed")
+    return repo
+
+
+def test_writable_profile_declares_write_and_version(tmp_path):
+    profile = GitShardAdapter("git", _repo(tmp_path, {"Home.md": "h"}), writable=True).profile()
+    assert profile.supports(Verb.WRITE)
+    assert profile.supports(Verb.VERSION)
+    profile.validate()  # PER_PAGE + WRITE is a consistent combination
+
+
+def test_write_creates_a_commit(tmp_path):
+    repo = _repo(tmp_path, {"Home.md": "old"})
+    adapter = GitShardAdapter("git", repo, writable=True)
+    before = _git(repo, "rev-list", "--count", "HEAD")
+    page = adapter.write("Home", "new body")
+    after = _git(repo, "rev-list", "--count", "HEAD")
+    assert int(after) == int(before) + 1  # one new commit
+    assert page.body == "new body"
+    assert page.envelope.source_rev == _git(repo, "rev-parse", "HEAD")  # page is at the new sha
+
+
+def test_write_advances_current_rev(tmp_path):
+    repo = _repo(tmp_path, {"Home.md": "old"})
+    adapter = GitShardAdapter("git", repo, writable=True)
+    rev_before = adapter.current_rev("Home")
+    adapter.write("Home", "changed")
+    assert adapter.current_rev("Home") != rev_before  # sha moved → drift detectable
+
+
+def test_write_new_key_tracks_it(tmp_path):
+    repo = _repo(tmp_path, {"Home.md": "h"})
+    adapter = GitShardAdapter("git", repo, writable=True)
+    adapter.write("docs/New", "fresh page")
+    assert "docs/New" in set(adapter.keys())
+    assert adapter.read("docs/New").body == "fresh page"
+
+
+def test_noop_write_creates_no_empty_commit(tmp_path):
+    repo = _repo(tmp_path, {"Home.md": "same"})
+    adapter = GitShardAdapter("git", repo, writable=True)
+    before = _git(repo, "rev-list", "--count", "HEAD")
+    adapter.write("Home", "same")  # identical body → nothing to commit
+    assert _git(repo, "rev-list", "--count", "HEAD") == before
+
+
+def test_current_rev_reflects_external_commit(tmp_path):
+    repo = _repo(tmp_path, {"Home.md": "h"})
+    adapter = GitShardAdapter("git", repo, writable=True)
+    rev = adapter.current_rev("Home")
+    # An out-of-band commit to the same path (another writer) moves the per-path sha.
+    (repo / "Home.md").write_text("externally edited", encoding="utf-8")
+    _git(repo, "add", "Home.md")
+    _git(repo, "commit", "-m", "external")
+    assert adapter.current_rev("Home") != rev
+
+
+def test_conformance_positive_write_probe_passes(tmp_path):
+    adapter = GitShardAdapter("git", _repo(tmp_path, {"Home.md": "body"}), writable=True)
+    report = run_conformance(adapter)
+    assert report.ok, report.diff()

tests/test_incremental_equivalence.py

@@ -0,0 +1,89 @@
+"""Tests for the indexed equivalence relation — blocking + verify (SHARD-WP-0011 T1)."""
+
+from itertools import combinations
+
+from shard_wiki.incremental import EquivalenceIndex, MinHasher, band_keys, jaccard, shingles
+from shard_wiki.incremental.equivalence import _fingerprint
+from shard_wiki.model import Identity, Page
+from shard_wiki.provenance import ProvenanceEnvelope
+
+
+def _page(shard, key, body):
+    return Page(
+        identity=Identity(shard, key),
+        body=body,
+        envelope=ProvenanceEnvelope(source_shard=shard),
+    )
+
+
+def _brute_force_groups(pages, threshold):
+    """Oracle: O(N²) verify of every pair, then connected components."""
+    parent = {p.identity: p.identity for p in pages}
+
+    def find(x):
+        while parent[x] != x:
+            parent[x] = parent[parent[x]]
+            x = parent[x]
+        return x
+
+    for p, q in combinations(pages, 2):
+        same_fp = _fingerprint(p.body) == _fingerprint(q.body)
+        sim = jaccard(shingles(p.body), shingles(q.body))
+        if same_fp or sim >= threshold:
+            parent[find(p.identity)] = find(q.identity)
+    comps = {}
+    for p in pages:
+        comps.setdefault(find(p.identity), set()).add(p.identity)
+    return {frozenset(v) for v in comps.values() if len(v) > 1}
+
+
+def test_minhash_lsh_buckets_near_duplicates_together():
+    hasher = MinHasher(num_perm=64)
+    base = "the quick brown fox jumps over the lazy dog near the river bank today"
+    near = base + " and then some"
+    far = "completely unrelated content about astrophysics and distant galaxies far"
+    b_base = set(band_keys(hasher.signature(shingles(base)), 32))
+    b_near = set(band_keys(hasher.signature(shingles(near)), 32))
+    b_far = set(band_keys(hasher.signature(shingles(far)), 32))
+    assert b_base & b_near  # near-duplicates share at least one band
+    assert not (b_base & b_far)  # unrelated pages do not
+
+
+def test_exact_duplicate_across_shards_is_equivalent():
+    idx = EquivalenceIndex()
+    idx.add(_page("A", "Foo", "identical body text here"))
+    idx.add(_page("B", "Bar", "identical body text here"))
+    assert idx.equivalent_to(Identity("A", "Foo")) == frozenset(
+        {Identity("A", "Foo"), Identity("B", "Bar")}
+    )
+
+
+def test_unrelated_pages_are_not_equivalent():
+    idx = EquivalenceIndex()
+    idx.add(_page("A", "Foo", "alpha beta gamma delta epsilon"))
+    idx.add(_page("B", "Bar", "nothing in common whatsoever entirely"))
+    assert idx.groups() == ()
+
+
+def test_curator_binding_forces_equivalence_regardless_of_content():
+    idx = EquivalenceIndex()
+    idx.add(_page("A", "Foo", "one thing"))
+    idx.add(_page("B", "Bar", "totally different"))
+    idx.bind(Identity("A", "Foo"), Identity("B", "Bar"))
+    assert idx.equivalent_to(Identity("A", "Foo")) == frozenset(
+        {Identity("A", "Foo"), Identity("B", "Bar")}
+    )
+
+
+def test_index_matches_brute_force_oracle():
+    threshold = 0.7
+    shared = "shared sentence one shared sentence two shared sentence three end"
+    pages = [
+        _page("A", "Doc1", shared),
+        _page("B", "Doc1copy", shared + " minor tail"),  # near-dup of A
+        _page("C", "Other", "a totally distinct page with no overlapping shingles at all here"),
+        _page("D", "Lonely", "yet another isolated document about unrelated subject matter alone"),
+    ]
+    idx = EquivalenceIndex(threshold=threshold)
+    idx.build(pages)
+    assert set(idx.groups()) == _brute_force_groups(pages, threshold)

tests/test_incremental_maintenance.py

@@ -0,0 +1,84 @@
+"""Incremental maintenance == rebuild, with retraction + propagation (SHARD-WP-0011 T2)."""
+
+from shard_wiki.incremental import EquivalenceIndex
+from shard_wiki.model import Identity, Page
+from shard_wiki.provenance import ProvenanceEnvelope
+
+
+def _page(shard, key, body):
+    return Page(
+        identity=Identity(shard, key),
+        body=body,
+        envelope=ProvenanceEnvelope(source_shard=shard),
+    )
+
+
+def _rebuilt(pages, curator=()):
+    idx = EquivalenceIndex()
+    idx.build(pages, curator)
+    return idx
+
+
+def _equal(a, b):
+    return a.edges() == b.edges() and set(a.groups()) == set(b.groups())
+
+
+def test_add_keeps_index_equal_to_rebuild():
+    pages = [_page("A", "Foo", "same content here"), _page("B", "Bar", "same content here")]
+    idx = EquivalenceIndex()
+    for p in pages:
+        idx.add(p)
+    assert _equal(idx, _rebuilt(pages))
+    assert idx.groups()  # the two collapse
+
+
+def test_remove_keeps_index_equal_to_rebuild():
+    pages = [
+        _page("A", "Foo", "same content here"),
+        _page("B", "Bar", "same content here"),
+        _page("C", "Baz", "unrelated isolated material entirely"),
+    ]
+    idx = _rebuilt(pages)
+    idx.remove(Identity("B", "Bar"))
+    assert _equal(idx, _rebuilt([pages[0], pages[2]]))
+
+
+def test_edit_into_new_bucket_retracts_stale_edge():
+    a = _page("A", "Foo", "shared identical body text")
+    b = _page("B", "Bar", "shared identical body text")
+    idx = _rebuilt([a, b])
+    assert idx.groups()  # A ≡ B initially
+    # Edit B to something completely different: it exits A's buckets, the edge is retracted.
+    b2 = _page("B", "Bar", "now totally divergent unrelated prose about nothing")
+    idx.update(b2)
+    assert idx.groups() == ()  # stale edge gone
+    assert _equal(idx, _rebuilt([a, b2]))
+
+
+def test_edit_into_equivalence_adds_edge():
+    a = _page("A", "Foo", "target body to converge on later")
+    b = _page("B", "Bar", "initially completely separate writing here")
+    idx = _rebuilt([a, b])
+    assert idx.groups() == ()
+    b2 = _page("B", "Bar", "target body to converge on later")  # now identical to A
+    idx.update(b2)
+    assert idx.equivalent_to(Identity("A", "Foo")) == frozenset(
+        {Identity("A", "Foo"), Identity("B", "Bar")}
+    )
+    assert _equal(idx, _rebuilt([a, b2]))
+
+
+def test_removing_connector_splits_a_chorus():
+    # Curator chain A—B—C (no direct A—C): one group of three.
+    a, b, c = (_page("A", "X", "aaa"), _page("B", "Y", "bbb"), _page("C", "Z", "ccc"))
+    idx = EquivalenceIndex()
+    for p in (a, b, c):
+        idx.add(p)
+    idx.bind(a.identity, b.identity)
+    idx.bind(b.identity, c.identity)
+    assert idx.equivalent_to(a.identity) == {a.identity, b.identity, c.identity}
+    # Removing the connector B retracts/propagates: the chorus splits.
+    idx.remove(b.identity)
+    assert idx.groups() == ()
+    chain = [(a.identity, b.identity), (b.identity, c.identity)]
+    assert _equal(idx, _rebuilt([a, c], curator=chain))

tests/test_incremental_verification.py

@@ -0,0 +1,89 @@
+"""Tests for I-2 verification — digest + consistency-checker (SHARD-WP-0011 T3)."""
+
+from shard_wiki.incremental import (
+    ConsistencyChecker,
+    EquivalenceIndex,
+    derived_digest,
+)
+from shard_wiki.model import Identity, Page
+from shard_wiki.provenance import ProvenanceEnvelope
+
+
+def _page(shard, key, body):
+    return Page(
+        identity=Identity(shard, key),
+        body=body,
+        envelope=ProvenanceEnvelope(source_shard=shard),
+    )
+
+
+def test_digest_is_stable_under_equivalent_event_orders():
+    pages = [
+        _page("A", "Foo", "shared body text here"),
+        _page("B", "Bar", "shared body text here"),
+        _page("C", "Baz", "an entirely separate unrelated document"),
+    ]
+    forward = EquivalenceIndex()
+    for p in pages:
+        forward.add(p)
+    reverse = EquivalenceIndex()
+    for p in reversed(pages):
+        reverse.add(p)
+    assert derived_digest(forward) == derived_digest(reverse)
+
+
+def test_clean_index_reports_healthy():
+    pages = [_page("A", "Foo", "same body"), _page("B", "Bar", "same body")]
+    idx = EquivalenceIndex()
+    idx.build(pages)
+    checker = ConsistencyChecker(idx, pages_fn := (lambda: pages))
+    report = checker.check_and_repair()
+    assert report.drifted is False and report.healthy is True
+    assert pages_fn()  # source unchanged
+
+
+def test_missed_delta_drift_is_detected_and_repaired():
+    a = _page("A", "Foo", "converging target body")
+    b = _page("B", "Bar", "initially unrelated separate text")
+    source = {"pages": [a, b]}
+    idx = EquivalenceIndex()
+    idx.build(source["pages"])
+    assert idx.groups() == ()  # not equivalent yet
+
+    # Source changes B to match A, but the index is never told (a missed delta → drift).
+    b2 = _page("B", "Bar", "converging target body")
+    source["pages"] = [a, b2]
+
+    checker = ConsistencyChecker(idx, lambda: source["pages"])
+    report = checker.check_and_repair()
+    assert report.drifted is True and report.repaired is True and report.healthy is True
+    # Self-healed: the index now reflects the equivalence.
+    assert idx.equivalent_to(Identity("A", "Foo")) == frozenset(
+        {Identity("A", "Foo"), Identity("B", "Bar")}
+    )
+
+
+def test_corrupted_internal_state_is_healed():
+    a = _page("A", "Foo", "identical content")
+    b = _page("B", "Bar", "identical content")
+    idx = EquivalenceIndex()
+    idx.build([a, b])
+    # Corrupt the derived tier directly: delete a true edge (simulated index corruption).
+    idx._content_edges.clear()
+    assert idx.groups() == ()  # corrupted away
+
+    checker = ConsistencyChecker(idx, lambda: [a, b])
+    report = checker.check_and_repair()
+    assert report.drifted is True and report.healthy is True
+    assert idx.groups()  # edge restored by scoped recompute
+
+
+def test_removed_source_page_is_reconciled():
+    a = _page("A", "Foo", "same body")
+    b = _page("B", "Bar", "same body")
+    idx = EquivalenceIndex()
+    idx.build([a, b])
+    checker = ConsistencyChecker(idx, lambda: [a])  # B vanished from source
+    report = checker.check_and_repair()
+    assert report.healthy is True
+    assert Identity("B", "Bar") not in idx.identities()

tests/test_incremental_wiring.py

@@ -0,0 +1,74 @@
+"""Wire the incremental tier behind InformationSpace views (SHARD-WP-0011 T4)."""
+
+from shard_wiki.adapters import FolderAdapter
+from shard_wiki.coordination import EventType
+from shard_wiki.model import Identity
+from shard_wiki.space import InformationSpace
+from shard_wiki.views import all_pages
+
+
+def _shard(tmp_path, name, files):
+    root = tmp_path / name
+    for rel, text in files.items():
+        p = root / rel
+        p.parent.mkdir(parents=True, exist_ok=True)
+        p.write_text(text, encoding="utf-8")
+    return FolderAdapter(name, root)
+
+
+def test_all_pages_via_index_matches_direct_fold(tmp_path):
+    space = InformationSpace("space")
+    space.attach(_shard(tmp_path, "wiki", {"Home.md": "welcome", "Guide.md": "the guide"}))
+    space.attach(_shard(tmp_path, "notes", {"Daily.md": "today"}))
+    # Routed-through-index result equals the direct fold-based computation (behaviour unchanged).
+    via_index = {(e.name, e.members) for e in space.all_pages()}
+    direct = {(e.name, e.members) for e in all_pages(space.union)}
+    assert via_index == direct
+
+
+def test_curator_binding_collapses_via_maintained_index(tmp_path):
+    space = InformationSpace("space")
+    space.attach(_shard(tmp_path, "a", {"Foo.md": "x"}))
+    space.attach(_shard(tmp_path, "b", {"Bar.md": "y"}))
+    space.log.append(
+        "space", EventType.BINDING_MADE, {"members": ["a:Foo", "b:Bar"]}
+    )
+    # The maintained index re-syncs curator edges live from the log fold.
+    collapsed = [e for e in space.all_pages() if len(e.members) == 2]
+    assert len(collapsed) == 1
+    assert set(collapsed[0].members) == {Identity("a", "Foo"), Identity("b", "Bar")}
+
+
+def test_content_duplicate_collapses_via_index(tmp_path):
+    space = InformationSpace("space")
+    space.attach(_shard(tmp_path, "a", {"Foo.md": "the very same body content here"}))
+    space.attach(_shard(tmp_path, "b", {"Bar.md": "the very same body content here"}))
+    dup = [e for e in space.all_pages() if len(e.members) == 2]
+    assert len(dup) == 1  # content equivalence detected by the maintained index
+    assert set(dup[0].members) == {Identity("a", "Foo"), Identity("b", "Bar")}
+
+
+def test_attach_invalidates_index(tmp_path):
+    space = InformationSpace("space")
+    space.attach(_shard(tmp_path, "a", {"Foo.md": "same body"}))
+    assert space.all_pages()  # builds the index (one page, no groups)
+    space.attach(_shard(tmp_path, "b", {"Bar.md": "same body"}))  # marks index stale
+    dup = [e for e in space.all_pages() if len(e.members) == 2]
+    assert len(dup) == 1  # rebuilt fallback picks up the new equivalent page
+
+
+def test_verify_index_reports_healthy_when_consistent(tmp_path):
+    space = InformationSpace("space")
+    space.attach(_shard(tmp_path, "a", {"Foo.md": "same body"}))
+    space.attach(_shard(tmp_path, "b", {"Bar.md": "same body"}))
+    space.all_pages()  # ensure built
+    report = space.verify_index()
+    assert report.healthy is True
+
+
+def test_reindex_is_an_explicit_fallback(tmp_path):
+    space = InformationSpace("space")
+    space.attach(_shard(tmp_path, "a", {"Foo.md": "content"}))
+    before = space.index.digest()
+    space.reindex()
+    assert space.index.digest() == before  # rebuild is deterministic

tests/test_views_allpages.py

@@ -0,0 +1,76 @@
+"""Tests for the AllPages + SiteMap enumeration views (SHARD-WP-0010 T4)."""
+
+from shard_wiki.adapters import FolderAdapter
+from shard_wiki.coordination import DecisionLog, EventType
+from shard_wiki.model import Identity
+from shard_wiki.union import UnionGraph
+from shard_wiki.views import all_pages, site_map
+
+
+def _shard(tmp_path, name, files):
+    root = tmp_path / name
+    for rel, text in files.items():
+        p = root / rel
+        p.parent.mkdir(parents=True, exist_ok=True)
+        p.write_text(text, encoding="utf-8")
+    return FolderAdapter(name, root)
+
+
+def test_all_pages_spans_shards(tmp_path):
+    u = UnionGraph("space")
+    u.attach(_shard(tmp_path, "shardA", {"A.md": "a"}))
+    u.attach(_shard(tmp_path, "shardB", {"B.md": "b"}))
+    names = {e.name for e in all_pages(u)}
+    assert names == {"A", "B"}
+
+
+def test_chorus_collapses_to_one_entry_with_divergence(tmp_path):
+    u = UnionGraph("space")
+    u.attach(_shard(tmp_path, "shardA", {"Home.md": "A home"}))
+    u.attach(_shard(tmp_path, "shardB", {"Home.md": "B home"}))
+    entries = all_pages(u)
+    home = [e for e in entries if e.name == "Home"]
+    assert len(home) == 1  # chorus → single entry
+    assert set(home[0].members) == {Identity("shardA", "Home"), Identity("shardB", "Home")}
+    assert home[0].diverges is True  # bodies differ — collapse acknowledged, not silent
+
+
+def test_chorus_same_body_does_not_diverge(tmp_path):
+    u = UnionGraph("space")
+    u.attach(_shard(tmp_path, "shardA", {"Home.md": "same"}))
+    u.attach(_shard(tmp_path, "shardB", {"Home.md": "same"}))
+    (home,) = [e for e in all_pages(u) if e.name == "Home"]
+    assert home.diverges is False
+
+
+def test_equivalence_binding_collapses_distinct_keys(tmp_path):
+    log = DecisionLog()
+    log.append(
+        "space", EventType.BINDING_MADE, {"members": ["shardA:Foo", "shardB:Bar"]}
+    )
+    u = UnionGraph("space", log=log)
+    u.attach(_shard(tmp_path, "shardA", {"Foo.md": "x"}))
+    u.attach(_shard(tmp_path, "shardB", {"Bar.md": "x"}))
+    pair = {Identity("shardA", "Foo"), Identity("shardB", "Bar")}
+    # The two bound identities fold into one entry (named by the min key, "Bar").
+    bound = [e for e in all_pages(u) if {*e.members} == pair]
+    assert len(bound) == 1
+    assert bound[0].name == "Bar"
+
+
+def test_sitemap_reflects_namespace_paths(tmp_path):
+    u = UnionGraph("space")
+    u.attach(
+        _shard(
+            tmp_path,
+            "shardA",
+            {"Home.md": "h", "docs/Guide.md": "g", "docs/api/Ref.md": "r"},
+        )
+    )
+    root = site_map(u)
+    # Top level: "Home" page directly, and a "docs" namespace.
+    assert any(p.key == "Home" for p in root.pages)
+    docs = next(c for c in root.children if c.name == "docs")
+    assert any(p.key == "docs/Guide" for p in docs.pages)
+    api = next(c for c in docs.children if c.name == "api")
+    assert any(p.key == "docs/api/Ref" for p in api.pages)

tests/test_views_backlinks.py

@@ -0,0 +1,51 @@
+"""Tests for the BackLinks derived view (SHARD-WP-0010 T2)."""
+
+from shard_wiki.adapters import FolderAdapter
+from shard_wiki.model import Identity
+from shard_wiki.union import UnionGraph
+from shard_wiki.views import build_backlinks
+
+
+def _shard(tmp_path, name, files):
+    root = tmp_path / name
+    for rel, text in files.items():
+        p = root / rel
+        p.parent.mkdir(parents=True, exist_ok=True)
+        p.write_text(text, encoding="utf-8")
+    return FolderAdapter(name, root)
+
+
+def test_link_yields_backlink_with_provenance(tmp_path):
+    u = UnionGraph("space")
+    u.attach(_shard(tmp_path, "shardA", {"A.md": "see [[B]]", "B.md": "target"}))
+    index = build_backlinks(u)
+    assert index.sources("B") == frozenset({Identity("shardA", "A")})
+    (bl,) = index.to("B")
+    assert bl.source_shard == "shardA"  # entry carries source provenance
+
+
+def test_red_links_create_no_backlinks(tmp_path):
+    u = UnionGraph("space")
+    u.attach(_shard(tmp_path, "shardA", {"A.md": "see [[Ghost]]"}))
+    index = build_backlinks(u)
+    assert index.to("Ghost") == ()  # unresolved target → no backlink
+    assert "Ghost" not in index.names()
+
+
+def test_chorus_target_aggregates_backlinks(tmp_path):
+    # "Home" exists in two shards (a chorus); links to it from anywhere aggregate under one name.
+    u = UnionGraph("space")
+    u.attach(_shard(tmp_path, "shardA", {"Home.md": "A home", "A.md": "[[Home]]"}))
+    u.attach(_shard(tmp_path, "shardB", {"Home.md": "B home", "B.md": "[[Home]]"}))
+    index = build_backlinks(u)
+    assert index.sources("Home") == frozenset(
+        {Identity("shardA", "A"), Identity("shardB", "B")}
+    )
+
+
+def test_backlinks_span_shards(tmp_path):
+    u = UnionGraph("space")
+    u.attach(_shard(tmp_path, "shardA", {"Index.md": "x"}))
+    u.attach(_shard(tmp_path, "shardB", {"B.md": "links [[Index]]"}))
+    index = build_backlinks(u)
+    assert index.sources("Index") == frozenset({Identity("shardB", "B")})

tests/test_views_integration.py

@@ -0,0 +1,52 @@
+"""Integration: derived views exposed on InformationSpace over two shards (SHARD-WP-0010 T5)."""
+
+from shard_wiki.adapters import FolderAdapter
+from shard_wiki.model import Identity
+from shard_wiki.space import InformationSpace
+
+
+def _shard(tmp_path, name, files):
+    root = tmp_path / name
+    for rel, text in files.items():
+        p = root / rel
+        p.parent.mkdir(parents=True, exist_ok=True)
+        p.write_text(text, encoding="utf-8")
+    return FolderAdapter(name, root)
+
+
+def _space(tmp_path):
+    space = InformationSpace("space")
+    space.attach(
+        _shard(tmp_path, "wiki", {"Home.md": "welcome, see [[Guide]]", "Guide.md": "the guide"})
+    )
+    space.attach(_shard(tmp_path, "notes", {"Daily.md": "today I read [[Guide]]"}))
+    return space
+
+
+def test_backlinks_across_two_shards(tmp_path):
+    space = _space(tmp_path)
+    sources = {bl.source for bl in space.backlinks("Guide")}
+    assert sources == {Identity("wiki", "Home"), Identity("notes", "Daily")}
+
+
+def test_all_pages_and_site_map_over_union(tmp_path):
+    space = _space(tmp_path)
+    names = {e.name for e in space.all_pages()}
+    assert names == {"Home", "Guide", "Daily"}
+    leaves = {p.key for p in space.site_map().pages}
+    assert {"Home", "Guide", "Daily"} <= leaves
+
+
+def test_recent_changes_includes_alias_and_edits(tmp_path):
+    space = _space(tmp_path)
+    space.alias("Start", "wiki:Home", actor="ana")
+    feed = space.recent_changes()
+    kinds = {e.kind for e in feed}
+    assert "alias" in kinds and "edit" in kinds
+    alias = next(e for e in feed if e.kind == "alias")
+    assert alias.source == "coordination" and alias.actor == "ana"
+
+
+def test_red_link_creates_no_backlink_via_space(tmp_path):
+    space = _space(tmp_path)
+    assert space.backlinks("Nonexistent") == ()

tests/test_views_links.py

@@ -0,0 +1,69 @@
+"""Tests for the wikilink + red-link model (SHARD-WP-0010 T1)."""
+
+from shard_wiki.adapters import FolderAdapter
+from shard_wiki.union import ResolutionKind, UnionGraph
+from shard_wiki.views import extract_links, resolve_links
+
+
+def _shard(tmp_path, name, files):
+    root = tmp_path / name
+    for rel, text in files.items():
+        p = root / rel
+        p.parent.mkdir(parents=True, exist_ok=True)
+        p.write_text(text, encoding="utf-8")
+    return FolderAdapter(name, root)
+
+
+def test_extracts_plain_and_labelled_links():
+    links = extract_links("See [[Home]] and [[Index|the index]].")
+    assert [(link.target, link.label, link.text) for link in links] == [
+        ("Home", None, "Home"),
+        ("Index", "the index", "the index"),
+    ]
+
+
+def test_links_carry_body_offsets_in_document_order():
+    body = "a [[One]] b [[Two]]"
+    links = extract_links(body)
+    assert [link.target for link in links] == ["One", "Two"]
+    s, e = links[0].span
+    assert body[s:e] == "[[One]]"
+
+
+def test_code_regions_are_not_scanned():
+    body = "real [[Home]]\n```\n[[NotALink]]\n```\ninline `[[AlsoNot]]` done"
+    targets = [link.target for link in extract_links(body)]
+    assert targets == ["Home"]
+
+
+def test_camelcase_off_by_default_then_opt_in():
+    body = "FrontPage links to [[Home]]"
+    assert [link.target for link in extract_links(body)] == ["Home"]  # CamelCase ignored
+    on = extract_links(body, camelcase=True)
+    assert {link.target for link in on} == {"FrontPage", "Home"}
+    assert next(link for link in on if link.target == "FrontPage").auto is True
+
+
+def test_camelcase_does_not_double_count_inside_explicit_link():
+    # [[FrontPage]] is one explicit link, not also a CamelCase auto-link.
+    links = extract_links("[[FrontPage]]", camelcase=True)
+    assert len(links) == 1
+    assert links[0].auto is False
+
+
+def test_resolve_links_distinguishes_link_from_red_link(tmp_path):
+    u = UnionGraph("space")
+    u.attach(_shard(tmp_path, "shardA", {"Home.md": "home"}))
+    resolved = resolve_links(u, "[[Home]] and [[Ghost]]")
+    by_target = {r.link.target: r for r in resolved}
+    assert by_target["Home"].resolution.kind is ResolutionKind.SINGLE
+    assert by_target["Home"].is_red_link is False
+    assert by_target["Ghost"].is_red_link is True  # unresolved → createable red-link
+
+
+def test_resolve_links_surfaces_chorus(tmp_path):
+    u = UnionGraph("space")
+    u.attach(_shard(tmp_path, "shardA", {"Home.md": "A"}))
+    u.attach(_shard(tmp_path, "shardB", {"Home.md": "B"}))
+    (resolved,) = resolve_links(u, "[[Home]]")
+    assert resolved.resolution.kind is ResolutionKind.CHORUS

tests/test_views_recentchanges.py

@@ -0,0 +1,67 @@
+"""Tests for the RecentChanges merged feed (SHARD-WP-0010 T3)."""
+
+import os
+from datetime import datetime, timezone
+
+from shard_wiki.adapters import FolderAdapter
+from shard_wiki.coordination import DecisionLog, EventType
+from shard_wiki.union import UnionGraph
+from shard_wiki.views import recent_changes
+
+
+def _shard(tmp_path, name, files, mtime=None):
+    root = tmp_path / name
+    for rel, text in files.items():
+        p = root / rel
+        p.parent.mkdir(parents=True, exist_ok=True)
+        p.write_text(text, encoding="utf-8")
+        if mtime is not None:
+            os.utime(p, (mtime, mtime))
+    return FolderAdapter(name, root)
+
+
+def test_edit_and_alias_both_appear_newest_first(tmp_path):
+    # Page edit signal pinned to an old mtime; the alias decision happens "now" → alias is newest.
+    old = datetime(2020, 1, 1, tzinfo=timezone.utc).timestamp()
+    u = UnionGraph("space")
+    u.attach(_shard(tmp_path, "shardA", {"Home.md": "home"}, mtime=old))
+    log = DecisionLog()
+    log.append("space", EventType.ALIAS_SET, {"alias": "Start", "target": "shardA:Home"})
+
+    feed = recent_changes(u, log, "space")
+    kinds = [e.kind for e in feed]
+    assert "edit" in kinds and "alias" in kinds
+    assert feed[0].kind == "alias"  # newest first
+    assert feed[-1].kind == "edit"
+    # Monotonic non-increasing by time.
+    assert all(feed[i].when >= feed[i + 1].when for i in range(len(feed) - 1))
+
+
+def test_per_shard_attribution_present(tmp_path):
+    u = UnionGraph("space")
+    u.attach(_shard(tmp_path, "shardA", {"A.md": "a"}))
+    u.attach(_shard(tmp_path, "shardB", {"B.md": "b"}))
+    feed = recent_changes(u, DecisionLog(), "space")
+    edits = {e.ref: e.source for e in feed if e.kind == "edit"}
+    assert edits["shardA:A"] == "shardA"
+    assert edits["shardB:B"] == "shardB"  # each edit attributed to its shard
+
+
+def test_coordination_entries_carry_actor_and_ref(tmp_path):
+    u = UnionGraph("space")
+    u.attach(_shard(tmp_path, "shardA", {"Doc.md": "x"}))
+    log = DecisionLog()
+    log.append(
+        "space", EventType.PAGE_FORKED, {"source": "shardA:Doc", "fork": "shardB:Doc"}, actor="ana"
+    )
+    fork = next(e for e in recent_changes(u, log, "space") if e.kind == "fork")
+    assert fork.source == "coordination"
+    assert fork.actor == "ana"
+    assert fork.ref == "shardA:Doc→shardB:Doc"
+
+
+def test_limit_truncates_to_newest(tmp_path):
+    u = UnionGraph("space")
+    u.attach(_shard(tmp_path, "shardA", {"A.md": "a", "B.md": "b", "C.md": "c"}))
+    feed = recent_changes(u, DecisionLog(), "space", limit=2)
+    assert len(feed) == 2

workplans/SHARD-WP-0001-yawex-requirements.md

@@ -2,7 +2,7 @@
 id: SHARD-WP-0001
 type: workplan
 title: "shard-wiki requirements from yawex prior art"
-domain: whynot
+domain: consumer
 repo: shard-wiki
 status: done
 owner: tegwick

workplans/SHARD-WP-0002-federation-architecture.md

@@ -2,7 +2,7 @@
 id: SHARD-WP-0002
 type: workplan
 title: "federation architecture design"
-domain: whynot
+domain: consumer
 repo: shard-wiki
 status: done
 owner: tegwick

workplans/SHARD-WP-0003-engine-dives-batch.md

@@ -2,7 +2,7 @@
 id: SHARD-WP-0003
 type: workplan
 title: "wiki-engine deep-dive batch (new-insight + git-forge + classic engines)"
-domain: whynot
+domain: consumer
 repo: shard-wiki
 status: done
 owner: tegwick

workplans/SHARD-WP-0004-computational-knowledge-systems.md

@@ -2,7 +2,7 @@
 id: SHARD-WP-0004
 type: workplan
 title: "computational / interactive-knowledge systems research"
-domain: whynot
+domain: consumer
 repo: shard-wiki
 status: done
 owner: tegwick

workplans/SHARD-WP-0005-architecture-hardening.md

@@ -2,7 +2,7 @@
 id: SHARD-WP-0005
 type: workplan
 title: "core architecture hardening (blueprint review fixes)"
-domain: whynot
+domain: consumer
 repo: shard-wiki
 status: done
 owner: tegwick

workplans/SHARD-WP-0006-architecture-hardening-2.md

@@ -2,7 +2,7 @@
 id: SHARD-WP-0006
 type: workplan
 title: "core architecture hardening II (round-2 review fixes)"
-domain: whynot
+domain: consumer
 repo: shard-wiki
 status: done
 owner: tegwick

workplans/SHARD-WP-0007-foundation-implementation.md

@@ -2,7 +2,7 @@
 id: SHARD-WP-0007
 type: workplan
 title: "foundation implementation — model, contract, decision log, union read"
-domain: whynot
+domain: consumer
 repo: shard-wiki
 status: done
 owner: tegwick

workplans/SHARD-WP-0008-write-path.md

@@ -2,7 +2,7 @@
 id: SHARD-WP-0008
 type: workplan
 title: "write path — overlay engine, writable adapter, apply-under-drift"
-domain: whynot
+domain: consumer
 repo: shard-wiki
 status: done
 owner: tegwick

workplans/SHARD-WP-0009-git-decision-log.md

@@ -2,7 +2,7 @@
 id: SHARD-WP-0009
 type: workplan
 title: "git-backed DecisionLog + per-space append authority"
-domain: whynot
+domain: consumer
 repo: shard-wiki
 status: done
 owner: tegwick

workplans/SHARD-WP-0010-derived-views.md

@@ -2,9 +2,9 @@
 id: SHARD-WP-0010
 type: workplan
 title: "derived views — wikilinks, BackLinks, RecentChanges, AllPages/SiteMap"
-domain: whynot
+domain: consumer
 repo: shard-wiki
-status: active
+status: done
 owner: tegwick
 topic_slug: whynot
 created: "2026-06-15"
@@ -36,7 +36,7 @@ later by SHARD-WP-0011) and carry provenance. Presentation stays out of core (L6
 
 ```task
 id: SHARD-WP-0010-T1
-status: todo
+status: done
 priority: high
 state_hub_task_id: "792660c3-9be9-4771-9f51-69d01f0c7f13"
 ```
@@ -51,7 +51,7 @@ red-link, CamelCase opt-in.
 
 ```task
 id: SHARD-WP-0010-T2
-status: todo
+status: done
 priority: high
 state_hub_task_id: "431a54c3-82b5-4b08-b3f0-762624d4c91d"
 ```
@@ -65,7 +65,7 @@ chorus pages aggregate.
 
 ```task
 id: SHARD-WP-0010-T3
-status: todo
+status: done
 priority: medium
 state_hub_task_id: "270c1c31-0445-42b9-9a49-92d32c298eb2"
 ```
@@ -79,7 +79,7 @@ alias both appear, newest-first; per-shard attribution present.
 
 ```task
 id: SHARD-WP-0010-T4
-status: todo
+status: done
 priority: low
 state_hub_task_id: "898ba43e-cdef-4ce8-9fa3-4ce60ebb4fdd"
 ```
@@ -92,7 +92,7 @@ collapses to one entry with divergence noted; sitemap reflects paths.
 
 ```task
 id: SHARD-WP-0010-T5
-status: todo
+status: done
 priority: medium
 state_hub_task_id: "7157544b-5d3b-45a2-ba5a-c32244c59323"
 ```

workplans/SHARD-WP-0011-incremental-union.md

@@ -2,9 +2,9 @@
 id: SHARD-WP-0011
 type: workplan
 title: "incremental union maintenance + equivalence index + I-2 verification"
-domain: whynot
+domain: consumer
 repo: shard-wiki
-status: active
+status: done
 owner: tegwick
 topic_slug: whynot
 created: "2026-06-15"
@@ -41,7 +41,7 @@ deployment is later.
 
 ```task
 id: SHARD-WP-0011-T1
-status: todo
+status: done
 priority: high
 state_hub_task_id: "842f480b-7b14-47cd-818b-012dbda9c187"
 ```
@@ -55,7 +55,7 @@ unrelated pages don't; verified edges match a brute-force oracle on a small corp
 
 ```task
 id: SHARD-WP-0011-T2
-status: todo
+status: done
 priority: high
 state_hub_task_id: "2da4e0b8-22cc-4ad1-a9aa-b5e991515d30"
 ```
@@ -70,7 +70,7 @@ stale edge.
 
 ```task
 id: SHARD-WP-0011-T3
-status: todo
+status: done
 priority: high
 state_hub_task_id: "b602ce31-ad9a-4c7f-b596-f039722373fc"
 ```
@@ -85,7 +85,7 @@ equivalent event orders.
 
 ```task
 id: SHARD-WP-0011-T4
-status: todo
+status: done
 priority: medium
 state_hub_task_id: "2f3d083c-0b2e-4b58-9e96-c0461c5eb089"
 ```

workplans/SHARD-WP-0012-git-shard-adapter.md

@@ -2,9 +2,9 @@
 id: SHARD-WP-0012
 type: workplan
 title: "second adapter — git-IS-store shard (contract validation on a new substrate)"
-domain: whynot
+domain: consumer
 repo: shard-wiki
-status: active
+status: done
 owner: tegwick
 topic_slug: whynot
 created: "2026-06-15"
@@ -40,7 +40,7 @@ merge beyond fast-forward (apply-under-drift refuse is enough, as in SHARD-WP-00
 
 ```task
 id: SHARD-WP-0012-T1
-status: todo
+status: done
 priority: high
 state_hub_task_id: "8a1c7c80-a0cc-4e02-a611-1f1fd7dec57b"
 ```
@@ -54,7 +54,7 @@ implication rules. Tests: read tracked files; profile validates; conformance rea
 
 ```task
 id: SHARD-WP-0012-T2
-status: todo
+status: done
 priority: high
 state_hub_task_id: "b47dfb86-46c1-4e97-a62f-377719499ff2"
 ```
@@ -68,7 +68,7 @@ changes after an external commit.
 
 ```task
 id: SHARD-WP-0012-T3
-status: todo
+status: done
 priority: medium
 state_hub_task_id: "4c895f42-671d-4948-8bdf-941fd85644bb"
 ```

workplans/SHARD-WP-0013-wiki-engine-prep.md

@@ -2,7 +2,7 @@
 id: SHARD-WP-0013
 type: workplan
 title: "wiki-engine prep — reuse-surface registration, UC-catalog systematization, WikiEngineCoreArchitecture"
-domain: whynot
+domain: consumer
 repo: shard-wiki
 status: done
 owner: tegwick

workplans/SHARD-WP-0014-engine-implementation.md

@@ -2,7 +2,7 @@
 id: SHARD-WP-0014
 type: workplan
 title: "wiki-engine implementation — kernel + typed-extension runtime + activation"
-domain: whynot
+domain: consumer
 repo: shard-wiki
 status: done
 owner: tegwick