feat(governance): add Crystal Projects v2 schema doc + issue import script

- governance/projects.md: canonical schema for Crystal Missions Project v2
  (custom fields, 6 canonical views, field assignment rules, intake policy).
- scripts/import-issues-to-project.py: import open issues from the 6 Crystal
  repos into the Project v2.
- Self-pin .crystal-governance.yaml bumped to v1.1.0.

Author: Malakof

.crystal-governance.yaml

@@ -1,5 +1,5 @@
 schema: crystal-governance-pin/v1
-governance_version: v1.0.2
+governance_version: v1.1.0
 source: Malakof/.github
 # Self-pin: Malakof/.github references its own current version.
 # Lets tooling (sync, governance-check) treat this repo like any other

README.md

@@ -3,7 +3,7 @@
 Single source of truth for Crystal team GitHub conventions:
 labels, issue/PR templates, naming, reusable workflows, agent skill.
 
-**Current version**: `v1.0.2` (see `.crystal-governance.yaml`).
+**Current version**: `v1.1.0` (see `.crystal-governance.yaml`).
 
 ## For humans
 
@@ -34,7 +34,7 @@ Add to the target repo:
 1. `.crystal-governance.yaml`:
    ```yaml
    schema: crystal-governance-pin/v1
-   governance_version: v1.0.2
+   governance_version: v1.1.0
    source: Malakof/.github
    ```
 
@@ -44,7 +44,7 @@ Add to the target repo:
    on: [push, pull_request]
    jobs:
      check:
-       uses: Malakof/.github/.github/workflows/governance-check.yml@v1.0.2
+       uses: Malakof/.github/.github/workflows/governance-check.yml@v1.1.0
    ```
 
 3. `.github/workflows/enforce-conventions.yml`:
@@ -53,7 +53,7 @@ Add to the target repo:
    on: [pull_request]
    jobs:
      enforce:
-       uses: Malakof/.github/.github/workflows/enforce-conventions.yml@v1.0.2
+       uses: Malakof/.github/.github/workflows/enforce-conventions.yml@v1.1.0
    ```
 
 4. Run label sync:

governance/projects.md

@@ -0,0 +1,122 @@
+# Crystal Projects v2 — canonical schema and views
+
+The Crystal mission portfolio is materialised as a single GitHub Projects v2
+board: **Crystal Missions** at https://github.com/users/Malakof/projects/1.
+This document is the canonical reference for the project's custom fields,
+canonical views, and policies.
+
+> **Status**: v1.0.2 — schema codified, views to be created via UI per
+> §2 below. Auto-sync from paperclip kernel will land in v1.1.0.
+
+## 1. Custom fields
+
+| Field | Type | Values |
+|---|---|---|
+| Title | text (default) | issue title |
+| Status | single-select (default) | Todo, In Progress, Done |
+| Mission code | text | `<REPO_PREFIX>-<TYPE>-<NUM>` (see [`README.md §3.1`](./README.md#31-mission-codes)) |
+| Priority | single-select | `p0`, `p1`, `p2`, `p3` |
+| Type | single-select | `epic`, `feature`, `bug`, `chore`, `docs`, `refactor`, `spike`, `test` |
+| Stream | single-select | `atlas`, `beacon`, `forge`, `compass` |
+| Runtime | single-select | `claude`, `codex`, `openhands` |
+| Release | text | `v1.0.0`, `v1.0-rc1`, `v2.0`, free form |
+| Repository | text (default) | auto from issue source |
+| Labels | text (default) | mirrored from the issue |
+| Linked pull requests | default | auto |
+| Milestone | default | auto |
+| Parent issue | default | auto |
+| Sub-issues progress | default | auto |
+
+Fields named in the canonical taxonomy mirror the labels documented in
+[`labels.yaml`](./labels.yaml). When the paperclip kernel projection is
+enabled (v1.1.0+), values are written by the kernel and **never edited
+manually**.
+
+## 2. Canonical views
+
+These six views are the supported lenses on the project. They are created
+manually via the GitHub UI today; v1.1.0 will provide a script to apply
+them via GraphQL.
+
+### View 1 — Operations Kanban (`Status`)
+
+- **Layout**: Board
+- **Group by**: Status
+- **Sort**: Priority ascending (p0 first)
+- **Columns shown**: Title, Mission code, Priority, Type, Repository,
+  Assignees
+- **Filter**: `is:open` AND (`Type` is not `epic`)
+- **Purpose**: daily ops view of in-flight work, excludes EPICs which
+  are tracked separately (see View 4).
+
+### View 2 — Priority × Stream (table)
+
+- **Layout**: Table
+- **Group by**: Priority
+- **Sort**: Stream then Mission code
+- **Columns shown**: Title, Mission code, Type, Stream, Runtime, Status,
+  Repository
+- **Filter**: `is:open`
+- **Purpose**: portfolio-wide capacity planning.
+
+### View 3 — Release Roadmap
+
+- **Layout**: Roadmap (or Table grouped by Release if Roadmap not available)
+- **Group by**: Release
+- **Filter**: `Release` is not empty
+- **Purpose**: cross-repo release readiness at a glance.
+
+### View 4 — EPIC tree (Hierarchy)
+
+- **Layout**: Table
+- **Filter**: `Type` is `epic` OR `Parent issue` is not empty
+- **Group by**: Parent issue (if available) else Repository
+- **Sort**: Priority
+- **Purpose**: track every EPIC and its native sub-issues progress.
+
+### View 5 — Per Runtime (load balance)
+
+- **Layout**: Board
+- **Group by**: Runtime
+- **Filter**: `is:open` AND `Status` is not `Done`
+- **Sort**: Priority then Mission code
+- **Purpose**: see what each runtime (Claude, Codex, OpenHands) is
+  currently working on.
+
+### View 6 — Triage queue
+
+- **Layout**: Table
+- **Filter**: label `status:triage` OR labels are empty (no `priority:p*`
+  AND no `type:*`)
+- **Sort**: created date descending
+- **Purpose**: catch issues that need product/impl triage.
+
+## 3. Field assignment rules
+
+| Source | Fields written | When |
+|---|---|---|
+| Issue creator (human) | Title, Type, Priority, Repository (auto) | At creation via ISSUE_TEMPLATE |
+| Paperclip kernel (v1.1.0+) | Mission code, Status, Stream, Runtime | On every mission stage transition |
+| Dark factory ingestion | Mission code (if not set), Type=epic on parents | At `epic-plan` ingestion |
+| GitHub auto | Linked pull requests, Sub-issues progress, Milestone, Parent issue | Continuously |
+
+**Don't edit manually** what the kernel projects (Mission code, Stream,
+Runtime, Status when kernel-driven). Manual edits will be overwritten on
+the next projection.
+
+## 4. Issue intake into the project
+
+For new repos onboarded after v1.0.2, ensure they are listed in
+`scripts/import-issues-to-project.py:REPOS`. New issues created from the
+provided ISSUE_TEMPLATE forms are automatically eligible for the project
+once added (manually for now via `gh project item-add`, automatically in
+v1.1.0).
+
+## 5. Coming in v1.1.0
+
+- Kernel-side projection: paperclip writes to Mission code / Stream /
+  Runtime / Status on every stage transition.
+- GraphQL view definitions: views above codified as JSON, applied via
+  `scripts/setup-project-views.py`.
+- `Crystal status` field added to extend the default `Status` with the
+  fine-grained `triage`, `ready`, `blocked`, `needs-human` states.

scripts/import-issues-to-project.py

@@ -0,0 +1,105 @@
+#!/usr/bin/env python3
+"""Import all open issues from Crystal repos into the Crystal Missions Project v2.
+
+Usage:
+    python scripts/import-issues-to-project.py [--state open|closed|all] [--dry-run]
+"""
+
+from __future__ import annotations
+
+import argparse
+import json
+import subprocess
+import sys
+import time
+
+REPOS = [
+    "Malakof/crystal-dark-factory-poc",
+    "Malakof/crystal-dark-factory-target-lab",
+    "Malakof/crystal-discord-bot",
+    "Malakof/crystal-specs",
+    "Malakof/crystal-assistant-ui-poc",
+    "Malakof/crystal-company",
+]
+
+PROJECT_NUMBER = 1
+PROJECT_OWNER = "Malakof"
+
+
+def gh_json(args: list[str]) -> object:
+    result = subprocess.run(["gh"] + args, capture_output=True, text=True, check=False)
+    if result.returncode != 0:
+        return {"_error": result.stderr.strip()[:200]}
+    if not result.stdout.strip():
+        return None
+    try:
+        return json.loads(result.stdout)
+    except json.JSONDecodeError:
+        return {"_raw": result.stdout[:200]}
+
+
+def list_issues(repo: str, state: str) -> list[dict]:
+    payload = gh_json([
+        "issue", "list",
+        "--repo", repo,
+        "--state", state,
+        "--json", "number,url,title",
+        "--limit", "500",
+    ]) or []
+    return payload if isinstance(payload, list) else []
+
+
+def add_to_project(url: str, dry_run: bool) -> str:
+    if dry_run:
+        return "would-add"
+    for attempt in range(3):
+        result = subprocess.run(
+            ["gh", "project", "item-add", str(PROJECT_NUMBER),
+             "--owner", PROJECT_OWNER, "--url", url],
+            capture_output=True, text=True, check=False,
+        )
+        if result.returncode == 0:
+            return "added"
+        err = result.stderr.strip()
+        if "already in the project" in err.lower() or "exists" in err.lower():
+            return "already-in-project"
+        if "504" in err or "502" in err or "Timeout" in err:
+            time.sleep(1.5 * (attempt + 1))
+            continue
+        return f"error:{err.splitlines()[0][:80]}"
+    return "max-retries"
+
+
+def main() -> int:
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--state", default="open", choices=["open", "closed", "all"])
+    parser.add_argument("--dry-run", action="store_true")
+    args = parser.parse_args()
+
+    summary: dict[str, dict[str, int]] = {}
+    total_issues = 0
+
+    for repo in REPOS:
+        issues = list_issues(repo, args.state)
+        repo_summary = {"total": len(issues), "added": 0, "already": 0, "error": 0, "would": 0}
+        for issue in issues:
+            status = add_to_project(issue["url"], args.dry_run)
+            if status == "added":
+                repo_summary["added"] += 1
+            elif status == "already-in-project":
+                repo_summary["already"] += 1
+            elif status == "would-add":
+                repo_summary["would"] += 1
+            else:
+                repo_summary["error"] += 1
+                print(f"[{repo} #{issue['number']}] {status}", file=sys.stderr)
+        summary[repo] = repo_summary
+        total_issues += repo_summary["total"]
+        print(f"  {repo}: {repo_summary}")
+
+    print(f"\nTotal: {total_issues} issues processed across {len(REPOS)} repos.")
+    return 0
+
+
+if __name__ == "__main__":
+    sys.exit(main())