feat(cloudflare): mirror method-page specs, drop the in-repo snapshot

.agents/skills/distilled-sdk/SKILL.md

@@ -50,7 +50,7 @@ this table is the whole decision.
 | Whole directories inside a huge git repo | `spec-repos/aws`, `spec-repos/azure` | blobless (`--filter=blob:none`) + `--no-checkout` clone, narrow sparse patterns from the tree, only then fetch blobs |
 | An index that enumerates many documents | `spec-repos/gcp` | crawl the discovery directory, write `_manifest.json` + one doc per entry |
 | A GraphQL endpoint | `spec-repos/railway`, `spec-repos/expo-eas` | introspection query → `schema.json` |
-| Only human docs (markdown/HTML) | nothing yet — read the cloudflare note below first | download + parse |
+| Only human docs (markdown/HTML) | `spec-repos/cloudflare` | crawl an index, download markdown, validate it is not HTML |
 
 A YAML spec does not have to be converted in the mirror: `spec-repos/coinbase`
 mirrors `openapi.yaml` verbatim and `packages/coinbase/scripts/convert.ts`
@@ -59,13 +59,12 @@ in the mirror only when the upstream is an endpoint rather than a file.
 
 **When the user says "like GitHub"** they mean a few files out of a big repo —
 raw download, no clone. **"Like cloudflare"** means the description has to be
-scraped out of documentation pages. `packages/cloudflare` is the cautionary
-example, not the model: its upstream moved to Astro and now serves HTML at
-every `index.md` URL, so `spec-mirror-cloudflare` is `blocked` in the manifest
-and the package runs off a stale snapshot. If you are about to scrape docs,
-first check whether the provider publishes a real machine-readable description
-somewhere — that is nearly always the better spec, and it is what unblocks
-cloudflare too (`cloudflare/api-schemas`).
+scraped out of documentation pages. Copy `spec-repos/cloudflare`: crawl an
+index, download markdown, and **reject HTML** (resource index pages on that
+host still serve the full HTML document at `index.md`). If you are about to
+scrape docs, first check whether the provider publishes a real
+machine-readable description somewhere — that is nearly always the better
+spec.
 
 Rules every `fetch-specs.ts` follows:
 
@@ -228,10 +227,8 @@ segment, which is why every package's declared path is
 A path that reads through some other layout resolves to a file the mirror
 does not have, and `resolveSpecPath` says so rather than failing later.
 
-Two packages are not wired to it. `aws` resolves its models directory inside
-`runGeneratorCli` rather than at the call site. `cloudflare` takes its spec
-root as a `--specs` flag and its mirror is `blocked` — see the stack README —
-so it is the one package still reading a stale in-repo snapshot.
+One package is not wired to it the usual way. `aws` resolves its models
+directory inside `runGeneratorCli` rather than at the call site.
 
 `fly-io` is the one package that reads more than its mirror: `specs/sprites`,
 `specs/mpg` and `specs/addons` are small hand-maintained documents committed

.gitmodules

@@ -13,6 +13,11 @@
 	url = https://github.com/distilled-mirror/spec-mirror-azure.git
 	shallow = true
 	ignore = dirty
+[submodule "packages/cloudflare/specs/spec-mirror-cloudflare"]
+	path = packages/cloudflare/specs/spec-mirror-cloudflare
+	url = https://github.com/distilled-mirror/spec-mirror-cloudflare.git
+	ignore = dirty
+	shallow = true
 [submodule "packages/coinbase/specs/spec-mirror-coinbase"]
 	path = packages/coinbase/specs/spec-mirror-coinbase
 	url = https://github.com/distilled-mirror/spec-mirror-coinbase.git

packages/cloudflare/package.json

@@ -500,12 +500,13 @@
   },
   "scripts": {
     "build": "tsc -b",
-    "download-api-docs": "bun scripts/download-api-docs.ts",
     "convert": "bun scripts/spec-to-smithy.ts",
     "spec-to-smithy": "bun scripts/spec-to-smithy.ts",
     "generate": "bun scripts/generate.ts",
     "typecheck": "tsc -p tsconfig.json",
-    "typecheck:strict": "tsc -p tsconfig.strict.json"
+    "typecheck:strict": "tsc -p tsconfig.strict.json",
+    "specs:fetch": "git submodule update --force --init --depth=1 specs/spec-mirror-cloudflare",
+    "specs:update": "git submodule update --remote --force --depth=1 specs/spec-mirror-cloudflare"
   },
   "dependencies": {
     "@distilled.cloud/core": "workspace:*"