Prepare MrScraper Claude plugin for community publishing

.claude-plugin/marketplace.json

@@ -3,20 +3,22 @@
   "name": "mrscraper-claude",
   "owner": {
     "name": "MrScraper",
+    "email": "support@mrscraper.com",
     "url": "https://github.com/mrscraper-com"
   },
-  "description": "MrScraper MCP skills and tools for Claude.",
+  "description": "Fetch and preserve page content, extract structured web data, discover sources, and work with saved MrScraper runs from Claude.",
   "plugins": [
     {
       "name": "mrscraper",
       "source": "./plugins/mrscraper",
-      "description": "Fetch page content, extract structured web data, discover sources through Google, and work with saved MrScraper runs.",
-      "version": "0.1.1",
+      "description": "Connect Claude to MrScraper for raw page fetching, managed extraction, Google discovery, and saved scraper workflows.",
       "author": {
-        "name": "MrScraper"
+        "name": "MrScraper",
+        "email": "support@mrscraper.com",
+        "url": "https://mrscraper.com"
       },
       "homepage": "https://docs.mrscraper.com/docs/getting-started/mcp-server",
-      "repository": "https://github.com/pray-mrscraper/mrscraper-claude-plugin",
+      "repository": "https://github.com/mrscraper-com/mrscraper-claude-plugin",
       "license": "MIT",
       "category": "productivity",
       "tags": [

.github/workflows/validate.yml

@@ -0,0 +1,31 @@
+name: Validate Claude plugin
+
+on:
+  pull_request:
+  push:
+    branches:
+      - main
+
+permissions:
+  contents: read
+
+jobs:
+  validate:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Check out repository
+        uses: actions/checkout@v4
+
+      - name: Set up Node.js
+        uses: actions/setup-node@v4
+        with:
+          node-version: 22
+
+      - name: Install Claude Code
+        run: npm install --global @anthropic-ai/claude-code@2.1.246
+
+      - name: Validate marketplace
+        run: claude plugin validate . --strict
+
+      - name: Validate plugin
+        run: claude plugin validate plugins/mrscraper --strict

CHANGELOG.md

@@ -0,0 +1,24 @@
+# Changelog
+
+All notable changes to the MrScraper Claude plugin are documented here. The
+project follows [Semantic Versioning](https://semver.org/).
+
+## 0.1.2 - 2026-08-26
+
+- Strengthened fetch-first routing, including reusable local extraction for
+  large same-layout page sets.
+- Added guidance for fetch Super Mode and managed extraction execution modes.
+- Added saved-rerun controls, exact result filters, and compact result polling.
+- Replaced placeholder targets with public Scrape This Site examples.
+- Added publication, data-handling, support, and security documentation.
+- Moved version ownership exclusively to the plugin manifest.
+
+## 0.1.1 - 2026-08-24
+
+- Moved the hosted MrScraper connection to Claude's native `.mcp.json` plugin
+  configuration.
+
+## 0.1.0 - 2026-08-24
+
+- Added the initial MrScraper Claude marketplace, MCP connection, and skill
+  pack.

PUBLISHING.md

@@ -0,0 +1,72 @@
+# Publishing MrScraper for Claude
+
+Use this checklist for a public MrScraper Claude plugin release.
+
+## Release gate
+
+- [ ] Merge only reviewed changes into `main`.
+- [ ] Keep the release version in
+      `plugins/mrscraper/.claude-plugin/plugin.json`; do not duplicate it in
+      the marketplace entry.
+- [ ] Record user-visible changes in `CHANGELOG.md`.
+- [ ] Confirm that the repository contains no credentials or private target
+      data.
+- [ ] Run both strict validations from a clean checkout:
+
+  ```bash
+  claude plugin validate . --strict
+  claude plugin validate plugins/mrscraper --strict
+  ```
+
+- [ ] Confirm the GitHub validation workflow passes.
+- [ ] Install the plugin from the public repository in a clean Claude profile,
+      complete OAuth, and confirm all seven MCP tools load.
+- [ ] Run the smoke-test prompt below.
+- [ ] Create the matching Git tag and GitHub release for an explicit semantic
+      version.
+
+## Submission details
+
+Keep these values ready for Anthropic's community marketplace form:
+
+| Field | Value |
+| --- | --- |
+| Plugin name | `mrscraper` |
+| Display name | `MrScraper` |
+| Repository | `https://github.com/mrscraper-com/mrscraper-claude-plugin` |
+| Plugin directory | `plugins/mrscraper` |
+| Homepage | `https://docs.mrscraper.com/docs/getting-started/mcp-server` |
+| Support | `support@mrscraper.com` |
+| Privacy policy | `https://mrscraper.com/privacy-policy` |
+| Acceptable use policy | `https://mrscraper.com/acceptable-use-policy` |
+| MCP endpoint | `https://mcp.mrscraper.com/mcp` |
+| Authentication | OAuth 2.1 browser authorization |
+
+Suggested summary:
+
+> Connect Claude to MrScraper for raw page fetching, managed structured
+> extraction, Google discovery, and reusable saved scraper workflows. The
+> included skills preserve source responses with a fetch-first workflow before
+> adding backend extraction when it is explicitly requested or clearly useful.
+
+The hosted service receives target URLs and tool inputs, including extraction
+instructions. Results are returned to Claude for the user's requested task. The
+OAuth connection can request `scrape:read`, `scrape:write`, and `account:read`
+access. The plugin contains no static credentials.
+
+Smoke-test prompt:
+
+```text
+Fetch https://www.scrapethissite.com/pages/simple/ and summarize the page.
+```
+
+## Submit to Anthropic
+
+After the release commit is on the public `main` branch, use Anthropic's
+[Console submission form](https://platform.claude.com/plugins/submit). A Team
+or Enterprise organization owner or directory manager can instead use the
+[Claude organization form](https://claude.ai/admin-settings/directory/submissions/plugins/new).
+
+Anthropic reviews third-party plugins for the `claude-community` marketplace.
+Approved entries are pinned to a repository commit, and the public catalog may
+take until its next nightly sync to show the plugin.

README.md

@@ -1,52 +1,134 @@
 # MrScraper for Claude
 
-This marketplace packages MrScraper's MCP-oriented skills together with the
-hosted MrScraper MCP server for Claude.
+MrScraper connects Claude to a hosted MCP server for fetching public web pages,
+managed structured extraction, Google discovery, saved scraper reruns, stored
+results, and account usage. The plugin also includes four focused skills that
+help Claude choose the right workflow and preserve raw page content.
 
-## Add the marketplace
+## What this plugin adds
 
-Paste this GitHub `owner/repo` value into Claude's **Add marketplace** dialog:
+- The hosted Streamable HTTP endpoint at `https://mcp.mrscraper.com/mcp`.
+- OAuth 2.1 authentication through Claude; no credential is stored in this
+  repository.
+- A fetch-first workflow that keeps raw page responses available for analysis,
+  verification, and follow-up transformations.
+- Managed extraction, site mapping, Google SERP discovery, saved reruns, and
+  stored-result tools when those capabilities are useful.
 
-```text
-pray-mrscraper/mrscraper-claude-plugin
-```
+## Requirements
 
-If the dialog specifically expects a Git repository URL, use
-`https://github.com/pray-mrscraper/mrscraper-claude-plugin.git`.
+- A current version of [Claude Code](https://code.claude.com/docs/en/setup).
+- A [MrScraper](https://app.mrscraper.com) account.
+- Permission to access and process the target content.
 
-From Claude Code CLI, add the marketplace and install the plugin:
+## Install
+
+Add this repository as a marketplace and install the plugin:
 
 ```bash
-claude plugin marketplace add pray-mrscraper/mrscraper-claude-plugin
+claude plugin marketplace add mrscraper-com/mrscraper-claude-plugin
 claude plugin install mrscraper@mrscraper-claude
 ```
 
-The plugin connects to `https://mcp.mrscraper.com/mcp`. When prompted, finish
-the OAuth authorization flow in your browser. Do not paste OAuth tokens or API
-keys into chat.
+In Claude's **Add marketplace** dialog, paste:
+
+```text
+mrscraper-com/mrscraper-claude-plugin
+```
+
+If the dialog expects a full Git repository URL, use:
+
+```text
+https://github.com/mrscraper-com/mrscraper-claude-plugin.git
+```
+
+Restart Claude or reload plugins after installation. Open `/mcp` if Claude asks
+you to authenticate, then complete the MrScraper OAuth flow in your browser.
+Never paste OAuth tokens or API keys into chat.
 
 ## Try it
 
 ```text
 Fetch https://www.scrapethissite.com/pages/simple/ and summarize the page.
 ```
 
-The included skills encourage `fetch` as the normal first step when a URL is
-known because it preserves the page response for flexible agent-led analysis.
-Use `scrape` when you specifically need backend structured extraction,
-pagination, schema-shaped output, or a reusable saved scraper.
+The connection should expose these MCP tools:
+
+| Tool | Purpose |
+| --- | --- |
+| `fetch` | Retrieve and preserve a known page's raw response. |
+| `scrape` | Run managed structured extraction or bounded site mapping. |
+| `serp` | Discover public pages through Google. |
+| `rerun` | Reuse a saved AI or manual scraper configuration. |
+| `results` | Browse and filter stored result records. |
+| `result` | Retrieve one stored result or poll an asynchronous run. |
+| `status` | Inspect subscription usage and request outcomes. |
+
+## Fetch-first routing
+
+When a public URL is already known, the skills direct Claude to fetch it first
+and treat the raw response as the source of truth. Claude can read, summarize,
+compare, or derive structured output locally without losing details to an
+early extraction prompt.
+
+This preference can still be faster for large same-layout sets. For roughly
+100 known pages, Claude can safely fetch pages concurrently, retain every raw
+response, and apply one reusable local extractor instead of requesting 100
+separate backend-LLM extractions. Use `scrape` when managed extraction is
+explicitly requested or has a clear benefit after the page structure and
+desired schema are understood.
 
-## Contents
+## Data and permissions
 
-- `plugins/mrscraper/.claude-plugin/plugin.json`: Claude plugin manifest
-- `plugins/mrscraper/skills/`: MCP-oriented MrScraper skill pack
-- `.claude-plugin/marketplace.json`: repository marketplace catalog
+The plugin sends MCP tool inputs—including target URLs and extraction
+instructions—to MrScraper's hosted service. Page responses and tool results are
+then available to Claude for the requested task. Use it only with public or
+otherwise authorized content, and follow the target site's requirements.
 
-## Development validation
+Review MrScraper's [MCP documentation](https://docs.mrscraper.com/docs/getting-started/mcp-server),
+[Privacy Policy](https://mrscraper.com/privacy-policy), and
+[Acceptable Use Policy](https://mrscraper.com/acceptable-use-policy) before use.
+
+## Update or remove
+
+```bash
+claude plugin marketplace update mrscraper-claude
+claude plugin update mrscraper@mrscraper-claude
+claude plugin uninstall mrscraper@mrscraper-claude
+```
+
+## Support and security
+
+- Product help: [MrScraper MCP documentation](https://docs.mrscraper.com/docs/getting-started/mcp-server)
+- Bugs and feature requests: [GitHub Issues](https://github.com/mrscraper-com/mrscraper-claude-plugin/issues)
+- Account help: [support@mrscraper.com](mailto:support@mrscraper.com)
+- Security reports: see [SECURITY.md](SECURITY.md)
+
+## Development
+
+The repository is both an installable Claude marketplace and the source of the
+`mrscraper` plugin:
+
+```text
+.claude-plugin/marketplace.json
+plugins/mrscraper/
+├── .claude-plugin/plugin.json
+├── .mcp.json
+└── skills/
+```
+
+Validate both manifests and all plugin components before a release:
 
 ```bash
 claude plugin validate . --strict
+claude plugin validate plugins/mrscraper --strict
 ```
 
-This repository is for marketplace testing before the equivalent packaging is
-integrated into the official MrScraper repositories.
+The plugin uses semantic versions from its plugin manifest. Bump that version
+for every release and record user-visible changes in [CHANGELOG.md](CHANGELOG.md).
+Use [PUBLISHING.md](PUBLISHING.md) for the release and Anthropic submission
+checklist.
+
+## License
+
+[MIT](LICENSE)

SECURITY.md

@@ -0,0 +1,12 @@
+# Security Policy
+
+Please report suspected vulnerabilities privately to
+[support@mrscraper.com](mailto:support@mrscraper.com). Include the affected
+component, reproduction steps, and potential impact. Do not open a public issue
+for an unpatched vulnerability or include credentials, tokens, customer data,
+or private target content in a report.
+
+The plugin repository contains no MrScraper credentials. Authentication is
+handled by Claude's OAuth 2.1 MCP connection. If a credential may have been
+exposed, revoke it through the relevant account or client immediately and then
+contact support.

plugins/mrscraper/.claude-plugin/plugin.json

@@ -2,14 +2,15 @@
   "$schema": "https://json.schemastore.org/claude-code-plugin-manifest.json",
   "name": "mrscraper",
   "displayName": "MrScraper",
-  "version": "0.1.1",
-  "description": "MrScraper MCP workflows for web fetching, structured extraction, Google discovery, and saved runs.",
+  "version": "0.1.2",
+  "description": "Connect Claude to MrScraper for raw page fetching, managed extraction, Google discovery, and saved scraper workflows.",
   "author": {
     "name": "MrScraper",
-    "url": "https://github.com/mrscraper-com"
+    "email": "support@mrscraper.com",
+    "url": "https://mrscraper.com"
   },
   "homepage": "https://docs.mrscraper.com/docs/getting-started/mcp-server",
-  "repository": "https://github.com/pray-mrscraper/mrscraper-claude-plugin",
+  "repository": "https://github.com/mrscraper-com/mrscraper-claude-plugin",
   "license": "MIT",
   "keywords": [
     "mrscraper",

plugins/mrscraper/README.md

@@ -0,0 +1,20 @@
+# MrScraper
+
+This Claude plugin connects to the hosted MrScraper MCP server and adds focused
+skills for raw page fetching, managed structured extraction, Google discovery,
+saved scraper reruns, stored results, and account usage.
+
+When a public URL is known, fetch is the normal first content-acquisition step.
+It preserves the page response so Claude can inspect every available detail,
+answer follow-up questions, and apply reusable local extraction logic. Even for
+roughly 100 same-layout pages, concurrent fetches followed by one local batch
+extractor can be faster and more complete than repeating backend-LLM extraction
+for every page. Use scrape when managed extraction is explicitly requested or
+still offers a clear benefit after the source and output schema are understood.
+
+The plugin connects to `https://mcp.mrscraper.com/mcp` and authenticates through
+OAuth 2.1. A MrScraper account is required. Do not paste OAuth tokens or API
+keys into Claude.
+
+See the repository [README](../../README.md) for installation, usage, data
+handling, update, and support instructions.