TypeSchema: fix versioned canonical resolution for cross-package base types

.claude/settings.json

@@ -0,0 +1,26 @@
+{
+  "hooks": {
+    "PreCompact": [
+      {
+        "hooks": [
+          {
+            "command": "bd prime",
+            "type": "command"
+          }
+        ],
+        "matcher": ""
+      }
+    ],
+    "SessionStart": [
+      {
+        "hooks": [
+          {
+            "command": "bd prime",
+            "type": "command"
+          }
+        ],
+        "matcher": ""
+      }
+    ]
+  }
+}

.gitignore

@@ -39,3 +39,8 @@ test-output
 /generated
 .typeschema-cache
 *.cpuprofile
+
+# Beads / Dolt files (added by bd init)
+.dolt/
+*.db
+.beads-credential-key

AGENTS.md

@@ -0,0 +1,84 @@
+# Agent Instructions
+
+This project uses **bd** (beads) for issue tracking. Run `bd prime` for full workflow context.
+
+## Quick Reference
+
+```bash
+bd ready              # Find available work
+bd show <id>          # View issue details
+bd update <id> --claim  # Claim work atomically
+bd close <id>         # Complete work
+bd dolt push          # Push beads data to remote
+```
+
+## Non-Interactive Shell Commands
+
+**ALWAYS use non-interactive flags** with file operations to avoid hanging on confirmation prompts.
+
+Shell commands like `cp`, `mv`, and `rm` may be aliased to include `-i` (interactive) mode on some systems, causing the agent to hang indefinitely waiting for y/n input.
+
+**Use these forms instead:**
+```bash
+# Force overwrite without prompting
+cp -f source dest           # NOT: cp source dest
+mv -f source dest           # NOT: mv source dest
+rm -f file                  # NOT: rm file
+
+# For recursive operations
+rm -rf directory            # NOT: rm -r directory
+cp -rf source dest          # NOT: cp -r source dest
+```
+
+**Other commands that may prompt:**
+- `scp` - use `-o BatchMode=yes` for non-interactive
+- `ssh` - use `-o BatchMode=yes` to fail instead of prompting
+- `apt-get` - use `-y` flag
+- `brew` - use `HOMEBREW_NO_AUTO_UPDATE=1` env var
+
+<!-- BEGIN BEADS INTEGRATION v:1 profile:minimal hash:ca08a54f -->
+## Beads Issue Tracker
+
+This project uses **bd (beads)** for issue tracking. Run `bd prime` to see full workflow context and commands.
+
+### Quick Reference
+
+```bash
+bd ready              # Find available work
+bd show <id>          # View issue details
+bd update <id> --claim  # Claim work
+bd close <id>         # Complete work
+```
+
+### Rules
+
+- Use `bd` for ALL task tracking — do NOT use TodoWrite, TaskCreate, or markdown TODO lists
+- Run `bd prime` for detailed command reference and session close protocol
+- Use `bd remember` for persistent knowledge — do NOT use MEMORY.md files
+
+## Session Completion
+
+**When ending a work session**, you MUST complete ALL steps below. Work is NOT complete until `git push` succeeds.
+
+**MANDATORY WORKFLOW:**
+
+1. **File issues for remaining work** - Create issues for anything that needs follow-up
+2. **Run quality gates** (if code changed) - Tests, linters, builds
+3. **Update issue status** - Close finished work, update in-progress items
+4. **PUSH TO REMOTE** - This is MANDATORY:
+   ```bash
+   git pull --rebase
+   bd dolt push
+   git push
+   git status  # MUST show "up to date with origin"
+   ```
+5. **Clean up** - Clear stashes, prune remote branches
+6. **Verify** - All changes committed AND pushed
+7. **Hand off** - Provide context for next session
+
+**CRITICAL RULES:**
+- Work is NOT complete until `git push` succeeds
+- NEVER stop before pushing - that leaves work stranded locally
+- NEVER say "ready to push when you are" - YOU must push
+- If push fails, resolve and retry until it succeeds
+<!-- END BEADS INTEGRATION -->

CHANGELOG.md

@@ -0,0 +1,12 @@
+# Changelog
+
+All notable changes to this project will be documented in this file.
+
+The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
+and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+
+## [Unreleased]
+
+### Fixed
+
+- **Versioned canonical resolution** for cross-package base types — packages with broken `.index.json` (e.g. `de.basisprofil.r4@1.5.4`) now fall back to direct directory scan via node_modules, enabling KBV profiles with versioned base type references to resolve correctly

CLAUDE.md

@@ -369,3 +369,51 @@ Detection uses `mkIsFamilyType(tsIndex)` which checks `schema.typeFamily.resourc
 - [Canonical Manager](https://github.com/atomic-ehr/canonical-manager)
 - [FHIR Schema](https://github.com/fhir-schema/fhir-schema)
 - [TypeSchema Spec](https://www.health-samurai.io/articles/type-schema-a-pragmatic-approach-to-build-fhir-sdk)
+
+
+<!-- BEGIN BEADS INTEGRATION v:1 profile:minimal hash:ca08a54f -->
+## Beads Issue Tracker
+
+This project uses **bd (beads)** for issue tracking. Run `bd prime` to see full workflow context and commands.
+
+### Quick Reference
+
+```bash
+bd ready              # Find available work
+bd show <id>          # View issue details
+bd update <id> --claim  # Claim work
+bd close <id>         # Complete work
+```
+
+### Rules
+
+- Use `bd` for ALL task tracking — do NOT use TodoWrite, TaskCreate, or markdown TODO lists
+- Run `bd prime` for detailed command reference and session close protocol
+- Use `bd remember` for persistent knowledge — do NOT use MEMORY.md files
+
+## Session Completion
+
+**When ending a work session**, you MUST complete ALL steps below. Work is NOT complete until `git push` succeeds.
+
+**MANDATORY WORKFLOW:**
+
+1. **File issues for remaining work** - Create issues for anything that needs follow-up
+2. **Run quality gates** (if code changed) - Tests, linters, builds
+3. **Update issue status** - Close finished work, update in-progress items
+4. **PUSH TO REMOTE** - This is MANDATORY:
+   ```bash
+   git pull --rebase
+   bd dolt push
+   git push
+   git status  # MUST show "up to date with origin"
+   ```
+5. **Clean up** - Clear stashes, prune remote branches
+6. **Verify** - All changes committed AND pushed
+7. **Hand off** - Provide context for next session
+
+**CRITICAL RULES:**
+- Work is NOT complete until `git push` succeeds
+- NEVER stop before pushing - that leaves work stranded locally
+- NEVER say "ready to push when you are" - YOU must push
+- If push fails, resolve and retry until it succeeds
+<!-- END BEADS INTEGRATION -->

COGNOVIS.md

@@ -0,0 +1,112 @@
+# cognovis/codegen
+
+This is a cognovis fork of [atomic-ehr/codegen](https://github.com/atomic-ehr/codegen) — the FHIR TypeScript/Python/C# code generator.
+
+## Current status: HIBERNATE
+
+**As of 2026-04-21:** The fork has no production-relevant fork-only code. All of our previous patches (the duplicate-`meta`-key fix, KBV on-the-fly workarounds) are merged upstream and ship in `@atomic-ehr/codegen@0.0.13`. Polaris / mira-adapters consumes `@atomic-ehr/codegen` directly from npm.
+
+The fork is retained for three narrow purposes:
+1. **Doc anchor** — this file and the roadmap in `.beads/`
+2. **PR staging** — clean `fix/` and `feat/` branches cut from `main` for upstream PRs
+3. **Future baking** — a place to develop and production-bake fork-only features before proposing them upstream
+
+When a fork-only feature is actively in development (see "Reactivation" below), we create a fresh consumer-snapshot branch on demand. We deliberately do **not** maintain a standing consumer branch, because that infrastructure (CI builds, committed `dist/`, force-push discipline) has real costs and only earns its keep when there is actual fork-only code to ship.
+
+## Reactivation criteria
+
+Leave hibernate mode when **all** of these hold:
+- A fork-only change is developed on `cognovis/next` (starts with the [FORK] epic `codegen-4cw`: scalar slice setters + input-type flattening, once unblocked)
+- Polaris / mira-adapters needs to production-bake that change before it lands upstream
+- The upstream PR is not yet merged (after merge, we return to hibernate and Polaris goes back to the npm release)
+
+Reactivation steps:
+1. Create a fresh `cognovis/mira-adapters` (or analogous) snapshot branch from `cognovis/next`
+2. Build `dist/`, commit it on the snapshot branch, push
+3. Polaris pins to the git URL: `github:cognovis/codegen#cognovis/mira-adapters`
+4. Automate rebuilds if the baking period is non-trivial (GHA on push to `cognovis/next`)
+5. After upstream merge and npm release, revert Polaris to the npm dep, archive the snapshot branch as `archive/<date>-<slug>` tag, delete the branch
+
+## Scope (when active)
+
+This fork covers **FHIR codegen extensions only** — vendor-neutral work that makes sense in a FHIR code generator. Aidbox-specific client code, persistence, validation-at-runtime etc. do **not** live here; they belong in `cognovis/aidbox-ts-sdk`.
+
+In-scope examples:
+- Bug fixes in the TypeScript profile writer
+- Scalar slice setters (`setBsnr("12345")` for slices with one primitive leaf after pattern-omit)
+- Input-type flattening (`Profile.fromInput({bsnr, ik})`)
+- Regression tests for profile patterns our IGs exercise
+
+Out-of-scope: anything that ties the generated output to a specific FHIR server (Aidbox, HAPI, etc.).
+
+## Release-gate philosophy
+
+Upstream `examples/on-the-fly/kbv-r4/` is the real release gate for anything that touches KBV: it pulls `kbv.ita.for@1.3.1` from Simplifier, generates types, and runs Runtime profile assertions. We rely on it instead of maintaining synthetic companion tests, because:
+- Synthetic tests that assert on emitted code structure (regex matches, snapshots) are brittle against legitimate generator refactors
+- The on-the-fly test catches the real-world failure modes (package index issues, dependency injection, KBV profile shapes) that synthetic tests miss
+- Running `bun test examples/on-the-fly/kbv-r4/` before any cognovis snapshot release is sufficient
+
+If we need offline/CI-resilient smoke coverage in the future, we add it then — not speculatively.
+
+## Branch model
+
+| Branch | Purpose | Sync |
+|---|---|---|
+| `main` | Pure mirror of [atomic-ehr/codegen `main`](https://github.com/atomic-ehr/codegen/tree/main). Never commit to this directly; always fast-forward from upstream. | `git fetch upstream && git reset --hard upstream/main && git push origin main` |
+| `cognovis/next` | Working / integrating branch. Fork-specific features and infra (this file, `.beads/`) live here on top of `main`. | Rebase onto `main` when syncing with upstream. |
+| `fix/<slug>`, `feat/<slug>` | Short-lived branches cut from pristine `main` for upstream PRs. Never base these on `cognovis/next` — keep them clean so the PR diff only shows the feature. | Delete after upstream merge or close. |
+| `cognovis/<consumer>` | **Not standing.** Created on-demand during reactivation (see above). | Created fresh each time; archived as tag after use. |
+
+### Current long-lived branches
+
+- `main` — upstream mirror at `a7a8dcf9` (atomic-ehr@0.0.13)
+- `cognovis/next` — upstream + fork infra (COGNOVIS.md, `.beads/`, CalVer bump)
+
+### Archived
+
+- `archive/2026-04-mira-adapters` — tag preserving the dismantled standing consumer branch from the pre-hibernate era. Restore with `git checkout -b cognovis/mira-adapters archive/2026-04-mira-adapters` if a similar reactivation is needed.
+
+## Upstream PR workflow
+
+1. Branch `fix/<slug>` or `feat/<slug>` from `main` — **not** `cognovis/next`. Upstream must see a clean, focused diff.
+2. Implement + test. Commit on the `fix/` branch with a conventional-commit message.
+3. Rebase `cognovis/next` on top to pick up the fix locally.
+4. `gh pr create --repo atomic-ehr/codegen --head cognovis:<branch>` to open the upstream PR.
+5. When upstream merges, delete the branch. The equivalent commit lands in `main` on the next upstream sync; `cognovis/next` rebases cleanly and our version of the commit drops out.
+
+If a change is inherently fork-only (e.g. infra, documentation, opinionated API surface we're not ready to propose upstream), document it in the commit message: `fork-only: <reason>`.
+
+## Upstream sync
+
+Cadence: on demand when (a) upstream ships a fix we want, (b) one of our open upstream PRs merges, or (c) periodically to avoid drift.
+
+```bash
+# 1. Sync main to upstream
+git checkout main && git fetch upstream && git reset --hard upstream/main && git push origin main
+
+# 2. Rebase cognovis/next onto updated main
+git checkout cognovis/next && git rebase main
+
+# 3. Re-run tests to catch regressions early
+bun test
+
+# 4. If a consumer-snapshot branch is live (reactivation phase), rebase it as well
+# Otherwise (hibernate), we are done.
+```
+
+See `.beads/` for the "Upstream sync runbook" bead with scripted tooling (in progress).
+
+## Project state & roadmap
+
+Tracked in `.beads/` (Dolt-backed). See `bd ready` for currently-actionable work.
+
+High-level roadmap:
+
+- **Hibernate maintenance** — keep `cognovis/next` rebased onto upstream; minimal ongoing cost.
+- **Fork-first feature: scalar slice setters + input-type flattening** (`codegen-4cw`). Currently blocked on fhir-de builder stabilization in Polaris. Reactivates the consumer-snapshot flow when development starts.
+- **Stabilise**: when/if atomic-ehr/codegen reaches 0.1.0 / 1.0.0 with a clear API contract, re-evaluate whether continuing to fork is still warranted.
+
+## Contact
+
+Technical: Malte Sussdorff (malte.sussdorff@cognovis.de)
+Upstream maintainer: [ryukzak](https://github.com/ryukzak) — responsive, open to PRs.

VERSION

@@ -0,0 +1 @@
+2026.04.0