Restructure journal into validated YAML with generated views

Move from prose-in-markdown to one YAML file per model under models/,
validated against a JSON Schema in CI. journal/*.md and INDEX.md are now
generated from the YAML by scripts/build.py — single source of truth.

Adds 11 new entries covering the April 2026 releases listed in the
@RoundtableSpace tweet (Claude Opus 4.7, DeepSeek V4, Xiaomi MiMi V2.5,
Qwen3.6 variants, GLM 5.1, Muse Spark, Grok 4.3 Beta, Kimi K2.6) as
stubs — enough to make them discoverable, sparse where sources weren't
available.

Also: CONTRIBUTING.md, PR template, GitHub Actions workflow that fails
if generated files drift from the YAML.

Author: OpenClaude

.github/pull_request_template.md

@@ -0,0 +1,18 @@
+<!-- Thanks for contributing. Fill in the relevant sections and delete the rest. -->
+
+## What this PR changes
+
+<!-- New model entry? Correction? Schema change? Tooling? -->
+
+## Checklist
+
+- [ ] Added or updated a YAML file under `models/`
+- [ ] Filename stem matches `model_id`
+- [ ] `release_date` is the actual public-availability date (not announcement if they differ)
+- [ ] At least one URL in `sources` (official announcement, docs, or model card)
+- [ ] Ran `python scripts/validate.py` locally — passes
+- [ ] Ran `python scripts/build.py` and committed the regenerated `journal/` and `INDEX.md`
+
+## Sources
+
+<!-- Paste the primary announcement URLs you used. -->

.github/workflows/validate.yml

@@ -0,0 +1,34 @@
+name: validate
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+
+jobs:
+  validate:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+          cache: pip
+
+      - name: Install dependencies
+        run: pip install pyyaml jsonschema
+
+      - name: Validate model YAML against schema
+        run: python scripts/validate.py
+
+      - name: Rebuild generated files
+        run: python scripts/build.py
+
+      - name: Fail if generated files are out of date
+        run: |
+          if ! git diff --exit-code journal/ INDEX.md; then
+            echo "::error::Generated files are out of date."
+            echo "Run 'python scripts/build.py' locally and commit the result."
+            exit 1
+          fi

.gitignore

@@ -0,0 +1,3 @@
+.venv/
+__pycache__/
+*.pyc

CONTRIBUTING.md

@@ -0,0 +1,47 @@
+# Contributing
+
+Thanks for helping keep the journal accurate.
+
+## Adding a new model
+
+1. Copy `models/TEMPLATE.yaml` to `models/<model_id>.yaml`. The filename
+   stem must equal the `model_id` field (the validator enforces this).
+2. Fill in as much as you know. Fields you're unsure about can stay `null`
+   — a sparse entry is fine, a wrong entry is not.
+3. Add at least one URL in `sources` pointing to a primary source (the
+   official announcement post, the provider's model card, or the API docs).
+4. Run the toolchain locally:
+   ```sh
+   python -m venv .venv
+   .venv/bin/pip install pyyaml jsonschema
+   .venv/bin/python scripts/validate.py
+   .venv/bin/python scripts/build.py
+   ```
+5. Commit the new YAML **and** the regenerated `journal/*.md` and
+   `INDEX.md`. CI will reject PRs where generated files drift from the
+   YAML.
+6. Open a PR. The template has a checklist.
+
+## Correcting an existing model
+
+Edit the YAML under `models/`. Don't edit `journal/*.md` or `INDEX.md` by
+hand — they're regenerated from the YAML.
+
+## Schema changes
+
+Changes to `schema/model.schema.json` are welcome but require:
+- Migration of every existing `models/*.yaml` in the same PR.
+- A note in the PR description explaining why the change is necessary.
+
+## What belongs here
+
+- New model releases from any provider.
+- Material updates to existing entries (pricing change, new modality,
+  context-window bump, deprecation).
+- Corrections.
+
+## What doesn't
+
+- Benchmarks and subjective reviews — this is a release record, not a
+  leaderboard.
+- Rumors or leaks without a primary source.

INDEX.md

@@ -0,0 +1,66 @@
+<!-- generated by scripts/build.py from models/*.yaml — do not edit -->
+
+# Model Release Index
+
+14 entries.
+
+| Date | Model | Provider | Context | Modalities | Status |
+|------|-------|----------|---------|------------|--------|
+| 2026-04-24 | [GPT-5.5](models/gpt-5.5.yaml) | OpenAI | 256K | text, code | stable |
+| 2026-04-24 | [Claude Sonnet 4.6](models/claude-sonnet-4-6.yaml) | Anthropic | 200K | text, image, pdf | stable |
+| 2026-04-24 | [Claude Opus 4.7](models/claude-opus-4-7.yaml) | Anthropic | 1M | text, image, pdf | stable |
+| 2026-04-24 | [Claude Opus 4.6](models/claude-opus-4-6.yaml) | Anthropic | 200K | text, image, pdf | stable |
+| 2026-04-01 | [Xiaomi MiMi V2.5](models/xiaomi-mimi-v2-5.yaml) | Xiaomi | — | text | stable |
+| 2026-04-01 | [Qwen3.6 Plus](models/qwen3.6-plus.yaml) | Alibaba | — | text | stable |
+| 2026-04-01 | [Qwen3.6 Max](models/qwen3.6-max.yaml) | Alibaba | — | text | stable |
+| 2026-04-01 | [Qwen3.6 35B A3B](models/qwen3.6-35b-a3b.yaml) | Alibaba | — | text | stable |
+| 2026-04-01 | [Qwen3.6 27B](models/qwen3.6-27b.yaml) | Alibaba | — | text | stable |
+| 2026-04-01 | [Muse Spark](models/muse-spark.yaml) | Unknown | — | text | stable |
+| 2026-04-01 | [Kimi K2.6](models/kimi-k2-6.yaml) | Moonshot AI | — | text | stable |
+| 2026-04-01 | [Grok 4.3 Beta](models/grok-4-3-beta.yaml) | xAI | — | text | preview |
+| 2026-04-01 | [GLM 5.1](models/glm-5-1.yaml) | Zhipu AI | — | text | stable |
+| 2026-04-01 | [DeepSeek V4](models/deepseek-v4.yaml) | DeepSeek | — | text, code | stable |
+
+## By provider
+
+### Alibaba
+
+- `qwen3.6-27b` — Qwen3.6 27B (2026-04-01)
+- `qwen3.6-35b-a3b` — Qwen3.6 35B A3B (2026-04-01)
+- `qwen3.6-max` — Qwen3.6 Max (2026-04-01)
+- `qwen3.6-plus` — Qwen3.6 Plus (2026-04-01)
+
+### Anthropic
+
+- `claude-opus-4-6` — Claude Opus 4.6 (2026-04-24)
+- `claude-opus-4-7` — Claude Opus 4.7 (2026-04-24)
+- `claude-sonnet-4-6` — Claude Sonnet 4.6 (2026-04-24)
+
+### DeepSeek
+
+- `deepseek-v4` — DeepSeek V4 (2026-04-01)
+
+### Moonshot AI
+
+- `kimi-k2-6` — Kimi K2.6 (2026-04-01)
+
+### OpenAI
+
+- `gpt-5.5` — GPT-5.5 (2026-04-24)
+
+### Unknown
+
+- `muse-spark` — Muse Spark (2026-04-01)
+
+### Xiaomi
+
+- `xiaomi-mimi-v2-5` — Xiaomi MiMi V2.5 (2026-04-01)
+
+### Zhipu AI
+
+- `glm-5-1` — GLM 5.1 (2026-04-01)
+
+### xAI
+
+- `grok-4-3-beta` — Grok 4.3 Beta (2026-04-01)
+

README.md

@@ -1,30 +1,62 @@
 # Model Release Journal
 
-An open, community-curated journal tracking model releases across the AI landscape.
+An open, community-curated record of AI model releases.
 
-## Why?
+- **Structured data**: one YAML file per model under [`models/`](models/).
+- **Validated**: every entry is checked against [`schema/model.schema.json`](schema/model.schema.json) in CI.
+- **Generated views**: [`INDEX.md`](INDEX.md) (all models, sortable by date or provider) and [`journal/`](journal/) (grouped by month) are rebuilt from the YAML — do not edit them by hand.
 
-Model releases happen constantly. Pricing changes, context windows expand, new capabilities drop. This journal exists to create a durable, diffable, community-editable record.
+See [`INDEX.md`](INDEX.md) for the current list.
 
-## Schema
+## Why
+
+Model releases happen constantly. Pricing shifts, context windows expand, naming conventions drift. Storing this as structured data — not prose — makes it diffable, queryable, and durable.
 
-Each entry tracks:
+## Schema
 
-| Field | Description |
-|-------|-------------|
-| `model_id` | The exact model identifier (e.g., `claude-sonnet-4-6`, `gpt-4.5`) |
-| `provider` | The serving provider (e.g., `Anthropic`, `OpenAI`, `xAI`) |
-| `release_date` | ISO 8601 date (`YYYY-MM-DD`) |
-| `context_window` | Context window in tokens (e.g., `200000`) |
-| `modalities` | Comma-separated list (e.g., `text, image, audio`) |
-| `x_post` | Link to the official X/Twitter announcement post |
-| `notes` | What changed, why it matters |
+Each `models/<model_id>.yaml` entry has these fields (see [`schema/model.schema.json`](schema/model.schema.json) for the authoritative definition):
+
+| Field | Required | Description |
+|-------|----------|-------------|
+| `model_id` | yes | Canonical slug, lowercase. Must match the filename stem. |
+| `display_name` | no | Human-readable name. |
+| `provider` | yes | Organization that serves the model. |
+| `release_date` | yes | ISO 8601 date of public availability. |
+| `context_window` | no | Input tokens (integer) or null. |
+| `output_tokens_max` | no | Max output tokens per response. |
+| `modalities` | no | Subset of `text, code, image, audio, video, pdf, embedding`. |
+| `status` | no | `preview`, `stable` (default), `deprecated`, `retired`. |
+| `supersedes` | no | `model_id` of the predecessor, if tracked here. |
+| `pricing` | no | `{ input_per_mtok, output_per_mtok, cached_input_per_mtok }`. |
+| `sources` | no | URLs of primary sources. |
+| `tags` | no | Free-form labels. |
+| `notes` | no | Short prose on what changed. |
 
 ## Contributing
 
-1. Add an entry to `journal/YYYY-MM.md`
-2. Keep entries in reverse chronological order (newest first)
-3. Include the X post link when available — it is the primary source
+See [`CONTRIBUTING.md`](CONTRIBUTING.md). Short version:
+
+```sh
+cp models/TEMPLATE.yaml models/<model_id>.yaml
+# edit fields
+python -m venv .venv && .venv/bin/pip install pyyaml jsonschema
+.venv/bin/python scripts/validate.py
+.venv/bin/python scripts/build.py
+git add models/ journal/ INDEX.md
+```
+
+## Repository layout
+
+```
+schema/model.schema.json   # JSON Schema for entries
+models/<model_id>.yaml     # one file per model (source of truth)
+models/TEMPLATE.yaml       # copy this to start a new entry
+scripts/validate.py        # schema + sanity checks (CI enforces)
+scripts/build.py           # regenerates journal/ and INDEX.md
+journal/YYYY-MM.md         # generated monthly view
+INDEX.md                   # generated global index
+.github/workflows/         # CI
+```
 
 ## License