feat: add hive MCP registry repo

hive-mcp-registry/.github/workflows/validate.yml

@@ -0,0 +1,47 @@
+name: Validate Registry
+
+on:
+  pull_request:
+
+jobs:
+  validate:
+    runs-on: ubuntu-latest
+    timeout-minutes: 1
+
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.11"
+
+      - name: Install validator dependencies
+        run: python -m pip install --disable-pip-version-check jsonschema requests
+
+      - name: Validate changed server manifests
+        shell: bash
+        run: |
+          set -euo pipefail
+          base_sha="${{ github.event.pull_request.base.sha }}"
+          head_sha="${{ github.event.pull_request.head.sha }}"
+          mapfile -t manifests < <(git diff --name-only "$base_sha" "$head_sha" -- registry/servers | grep '/manifest\.json$' | sort -u || true)
+
+          if [ "${#manifests[@]}" -eq 0 ]; then
+            echo "No server manifests changed."
+          else
+            for manifest in "${manifests[@]}"; do
+              python scripts/validate.py "$manifest"
+            done
+          fi
+
+      - name: Validate bundled examples
+        shell: bash
+        run: |
+          set -euo pipefail
+          for manifest in registry/_examples/*/manifest.json; do
+            python scripts/validate.py "$manifest"
+          done

hive-mcp-registry/CONTRIBUTING.md

@@ -0,0 +1,183 @@
+# Contributing to hive-mcp-registry
+
+## 5-Minute Quickstart
+
+1. Fork the repository and clone your fork locally.
+2. Copy `registry/_template/` to `registry/servers/<your-server>/`.
+3. Update `manifest.json` with your real package, transport, credentials, and docs.
+4. Add a sibling `README.md` that explains install, authentication, and a quick usage example.
+5. Run:
+
+```bash
+python scripts/validate.py registry/servers/<your-server>/manifest.json
+python scripts/build_index.py
+```
+
+6. Commit your changes and open a pull request.
+
+## Manifest Reference
+
+| Field | Type | Required | Description |
+| --- | --- | --- | --- |
+| `$schema` | `string` | No | Canonical schema URL for editor tooling and validation. |
+| `name` | `string` | Yes | Registry identifier. Use lowercase letters, numbers, and hyphens only. |
+| `display_name` | `string` | Yes | Human-friendly title shown in search results and UIs. |
+| `version` | `string` | Yes | Server version in semver form, such as `1.2.0`. |
+| `description` | `string` | Yes | Short summary of the server's purpose. |
+| `author` | `object` | Yes | Original creator metadata with `name`, `github`, and `url`. |
+| `maintainer` | `object` | Yes | Current maintainer metadata with `github` and `email`. |
+| `repository` | `string` | Yes | Source repository URL. |
+| `license` | `string` | Yes | License label or SPDX-like identifier. |
+| `status` | `string` | Yes | One of `official`, `verified`, or `community`. |
+| `docs_url` | `string` | No | Canonical documentation URL for install and auth setup. |
+| `supported_os` | `string[]` | No | Supported OS values: `linux`, `macos`, `windows`. |
+| `deprecated` | `boolean` | No | Whether the entry should no longer be installed. |
+| `deprecated_by` | `string` | No | Registry name of the preferred replacement. |
+| `transport` | `object` | Yes | Transport metadata with `supported` and `default`. |
+| `install.pip` | `string \| null` | Yes | PyPI package name, or `null` if unavailable. |
+| `install.docker` | `string \| null` | Yes | Docker image reference, or `null` if unavailable. |
+| `install.npm` | `string \| null` | Yes | npm package name, or `null` if unavailable. |
+| `stdio` | `object` | No | Launch config with `command` and `args`. |
+| `http` | `object` | No | Launch config with `default_port`, `health_path`, `command`, and `args`. |
+| `unix` | `object` | No | Launch config with `socket_template`, `command`, and `args`. |
+| `sse` | `object` | No | Launch config with `url_template`, `command`, and `args`. |
+| `tools` | `object[]` | Yes | Tool definitions with `name` and `description`. Tool names must be snake_case. |
+| `credentials` | `object[]` | Yes | Credential definitions with `id`, `env_var`, `description`, optional `help_url`, and `required`. |
+| `tags` | `string[]` | Yes | Search taxonomy tags. Every tag must come from the allowed list below. |
+| `categories` | `string[]` | Yes | High-level category labels used by clients. |
+| `mcp_protocol_version` | `string` | Yes | MCP protocol version date, such as `2024-11-05`. |
+| `hive.min_version` | `string` | If `hive` present | Minimum Hive version supported. |
+| `hive.max_version` | `string \| null` | If `hive` present | Maximum Hive version supported, or `null`. |
+| `hive.profiles` | `string[]` | If `hive` present | Hive profiles where this server is a strong fit. |
+| `hive.tool_namespace` | `string` | If `hive` present | Preferred Hive namespace prefix for tools. |
+| `hive.example_agent` | `string` | If `hive` present | URL to an example Hive agent using the server. |
+
+`_comment` fields are allowed anywhere in the manifest and are ignored by tooling. Use them sparingly to make templates clearer.
+
+## Transport Examples
+
+### stdio
+
+Use stdio when the client launches the MCP server as a subprocess.
+
+```json
+{
+  "transport": { "supported": ["stdio"], "default": "stdio" },
+  "stdio": {
+    "command": "uvx",
+    "args": ["my-server", "--stdio"]
+  }
+}
+```
+
+### http
+
+Use HTTP when the server exposes a local or remote web service and can report health.
+
+```json
+{
+  "transport": { "supported": ["http"], "default": "http" },
+  "http": {
+    "default_port": 4010,
+    "health_path": "/health",
+    "command": "uvx",
+    "args": ["my-server", "--http", "--port", "{port}"]
+  }
+}
+```
+
+### unix
+
+Use Unix sockets for low-overhead local communication on Linux or macOS.
+
+```json
+{
+  "transport": { "supported": ["unix"], "default": "unix" },
+  "unix": {
+    "socket_template": "/tmp/mcp-{name}.sock",
+    "command": "uvx",
+    "args": ["my-server", "--unix", "{socket_path}"]
+  }
+}
+```
+
+### sse
+
+Use SSE when the server already exposes a long-lived stream endpoint.
+
+```json
+{
+  "transport": { "supported": ["sse"], "default": "sse" },
+  "sse": {
+    "url_template": "http://localhost:{port}/sse",
+    "command": "npx",
+    "args": ["@scope/my-server", "--sse", "--port", "{port}"]
+  }
+}
+```
+
+## Tag Taxonomy
+
+| Tag | Use when the server is primarily about... |
+| --- | --- |
+| `project-management` | Planning, task orchestration, boards, or delivery tracking |
+| `communication` | Chat, messaging, notifications, or collaboration workflows |
+| `crm` | Customer records, deal pipelines, or account management |
+| `finance` | Accounting, budgets, billing, procurement, or payments |
+| `developer-tools` | Engineering workflows, source control, build systems, or local tooling |
+| `productivity` | General business operations and task execution |
+| `data` | Data retrieval, transformation, enrichment, or warehousing |
+| `atlassian` | Jira, Confluence, or Atlassian ecosystem integrations |
+| `google` | Google Workspace or Google Cloud ecosystem integrations |
+| `microsoft` | Microsoft 365, Azure, or Microsoft platform integrations |
+| `aws` | Amazon Web Services infrastructure or data services |
+| `github` | GitHub repositories, issues, or pull requests |
+| `issue-tracking` | Tickets, bugs, defects, or incident queues |
+| `calendar` | Scheduling, events, attendees, or meeting coordination |
+| `email` | Mailboxes, drafts, sending, or inbox triage |
+| `database` | Relational or document database access |
+| `search` | Web search, enterprise search, or retrieval APIs |
+| `ai` | LLM-adjacent workflows, model providers, or AI-native services |
+| `analytics` | Reporting, dashboards, BI, or event analysis |
+| `storage` | File storage, buckets, objects, or knowledge repositories |
+| `monitoring` | Observability, incidents, uptime, or alerting |
+| `security` | Security testing, auth, secrets, posture, or compliance |
+| `ecommerce` | Orders, products, storefronts, or fulfillment |
+| `hr` | Recruiting, people ops, interviews, or employee systems |
+| `marketing` | Campaigns, content, attribution, or lead generation |
+
+## CI Checks
+
+The GitHub Action validates:
+
+1. JSON syntax and schema compliance.
+2. Registry naming rules and duplicate names.
+3. PyPI package existence when `install.pip` is set.
+4. snake_case tool names and UPPER_SNAKE_CASE credential env vars.
+5. Allowed tag taxonomy usage.
+6. Example manifests under `registry/_examples/`.
+
+Common fixes:
+
+- `✗ [SCHEMA]`: compare your manifest against `schema/manifest.schema.json`.
+- `✗ [NAME FORMAT]`: rename `name` to lowercase letters, numbers, and hyphens only.
+- `✗ [PYPI]`: publish the package first, correct the package name, or set `install.pip` to `null`.
+- `✗ [TAGS]`: swap any unsupported tag for one from the taxonomy table above.
+- `✗ [README]`: add a sibling `README.md` for real submissions under `registry/servers/`.
+
+## Review Process
+
+Maintainers aim to review straightforward submissions within 3-5 business days.
+
+Reviewers look for:
+
+- Accurate install instructions and transport metadata.
+- Clear credential guidance and working docs.
+- Tool descriptions that are specific enough for discovery UIs.
+- Appropriate trust tier selection.
+
+Trust tier promotion usually works like this:
+
+- New third-party submissions start as `community`.
+- Stable entries with reproducible install and auth steps can move to `verified`.
+- `official` is reserved for Aden Hive-maintained or explicitly endorsed servers.