Point at the live listing, and correct the 1.0.0 claims

CHANGELOG.md

@@ -2,7 +2,7 @@
 
 Every released version has a [GitHub Release](https://github.com/devpeer-net/website-blocker/releases)
 carrying the exact `.zip` that is uploaded to the Chrome Web Store, its SHA-256, a build
-provenance attestation, and a tree digest you can regenerate from an installed copy to
+provenance attestation, and a content digest you can regenerate from an installed copy to
 check that the store is serving that same build. See [docs/RELEASING.md](docs/RELEASING.md)
 for how a release is cut and how to verify one.
 
@@ -12,15 +12,10 @@ full semver, because Chrome rejects pre-release suffixes such as `1.1.0-rc.1`.
 
 ## [Unreleased]
 
-### Changed
-
-- The extension name and store summary are the final listing copy:
-  `Website Blocker - Block Distracting Sites, Private & Simple`. The toolbar tooltip stays
-  plain `Website Blocker`.
-
 ## [1.0.0] — 2026-08-08
 
-First release.
+First release. Published to the Chrome Web Store as
+`Website Blocker - Block Distracting Sites, Private & Simple`.
 
 ### Added
 

README.md

@@ -19,17 +19,23 @@ switching blocking off — and keeps everything else frictionless.
 
 ## Install
 
-Not yet on the Chrome Web Store. Take the zip from the
-[latest release](https://github.com/devpeer-net/website-blocker/releases/latest) — or build
-it yourself:
+**[Website Blocker on the Chrome Web Store](https://chromewebstore.google.com/detail/website-blocker-block-dis/ilkclodiadkjmnclbkghbnbhnplnfhia)**
+
+You do not have to take the store's word for what it is serving. Every release publishes a
+content digest, and [`scripts/verify-install.mjs`](scripts/verify-install.mjs) checks the
+installed extension against it — see [Releases](#releases).
+
+Or run it unpacked, from the zip on the
+[latest release](https://github.com/devpeer-net/website-blocker/releases/latest) or from
+source:
 
 ```bash
 pnpm install
 pnpm build
 ```
 
-Either way, open `chrome://extensions`, turn on **Developer mode**, choose **Load
-unpacked**, and select the `dist/` folder (or the unzipped release).
+Then open `chrome://extensions`, turn on **Developer mode**, choose **Load unpacked**, and
+select the `dist/` folder (or the unzipped release).
 
 Two settings are worth changing right after installing:
 

legal/README.md

@@ -6,9 +6,15 @@ as a single public object on S3:
     https://website-blocker-extension.s3.us-east-1.amazonaws.com/privacy.html
 
 `bucket-policy.json` is the policy applied to that bucket. It grants anonymous
-`s3:GetObject` on exactly that one key — the bucket is not listable and no other object is
-readable — and denies everything over plain HTTP, so the policy the listing cites cannot be
-tampered with in transit.
+`s3:GetObject` on exactly that one key; the bucket is not listable and no other object is
+readable.
+
+There is deliberately no deny-non-TLS rule. The usual reason for one is confidentiality or
+integrity of something worth intercepting, and this is a public document whose entire
+content is already world-readable. The URL given to the Web Store is `https://`, so that is
+what anyone actually follows. Blocking plain HTTP would buy a defence against an attacker
+who can already MITM the connection but only wants to reword a privacy policy — which is
+not a threat worth a rule that can fail closed later.
 
     aws s3api put-bucket-policy --bucket website-blocker-extension \
       --policy file://legal/bucket-policy.json

legal/bucket-policy.json

@@ -7,21 +7,6 @@
       "Principal": "*",
       "Action": "s3:GetObject",
       "Resource": "arn:aws:s3:::website-blocker-extension/privacy.html"
-    },
-    {
-      "Sid": "DenyInsecureTransport",
-      "Effect": "Deny",
-      "Principal": "*",
-      "Action": "s3:*",
-      "Resource": [
-        "arn:aws:s3:::website-blocker-extension",
-        "arn:aws:s3:::website-blocker-extension/*"
-      ],
-      "Condition": {
-        "Bool": {
-          "aws:SecureTransport": "false"
-        }
-      }
     }
   ]
 }

scripts/verify-install.mjs

@@ -187,7 +187,9 @@ function report(label, digest, count, extra = '') {
   console.log(`${label.padEnd(10)} ${digest}  (${count} files${extra})`)
 }
 
-if (import.meta.url === pathToFileURL(process.argv[1]).href) main()
+// argv[1] is undefined under `node -e` and in some loaders, and pathToFileURL throws on
+// undefined — which would crash package.mjs's import rather than skipping the CLI.
+if (process.argv[1] && import.meta.url === pathToFileURL(process.argv[1]).href) main()
 
 function main() {
   const options = {}