Rust MCP server and CLI for Apprise notification fan-out across dozens of delivery backends, with stdio/HTTP transports, auth, and plugin packaging.
claude mcp add rapprise -- npx -y @dinglebear/rapprise{
"mcpServers": {
"rapprise": {
"command": "npx",
"args": ["-y", "@dinglebear/rapprise"],
"env": {
"APPRISE_MCP_HOST": "<apprise_mcp_host>",
"APPRISE_URL": "<apprise_url>",
"APPRISE_TOKEN": "<apprise_token>"
}
}
}
}APPRISE_MCP_HOSTAPPRISE_URLAPPRISE_TOKENMCP Servers overview
# apprise-rmcp
Apprise notifications over MCP and CLI with authenticated stdio and HTTP transports.
It exposes one MCP tool, `apprise`, plus the `rapprise` CLI. Agents can send
tagged notifications through a preconfigured Apprise server, run one-off
Apprise URL sends, and check upstream health through stdio MCP, Streamable HTTP
MCP, or direct shell commands.
**30-second path:** run `npx -y @dinglebear/rapprise health --json` -> start loopback
HTTP with `APPRISE_MCP_HOST=127.0.0.1 npx -y @dinglebear/rapprise serve` -> call
`tools/call` with `{"action":"health"}`.
**Status:** operational RMCP upstream-client server. Write-capable; notification
sends are intentional side effects. HTTP MCP supports loopback dev mode, static
bearer tokens, and Google OAuth through `lab-auth`.
**Not for:** replacing Apprise API, storing notification destinations in this
repo, scheduling reminders, building a generic webhook relay, multi-tenant
isolation, or passing upstream Apprise bearer tokens through MCP tool arguments.
## Contents
- [Naming](#naming)
- [Capabilities And Boundaries](#capabilities-and-boundaries)
- [Install](#install)
- [Quickstart](#quickstart)
- [Client Configuration](#client-configuration)
- [Runtime Surfaces](#runtime-surfaces)
- [MCP Tool Reference](#mcp-tool-reference)
- [CLI Reference](#cli-reference)
- [Configuration](#configuration)
- [Authentication](#authentication)
- [Safety And Trust Model](#safety-and-trust-model)
- [Architecture](#architecture)
- [Distribution Contract](#distribution-contract)
- [Development](#development)
- [Verification](#verification)
- [Deployment](#deployment)
- [Troubleshooting](#troubleshooting)
- [Related Servers](#related-servers)
- [Documentation](#documentation)
- [License](#license)
## Naming
| Surface | This repo |
|---|---|
| Repository | `apprise-rmcp` |
| Rust crate | `apprise-mcp` |
| Binary / CLI | `rapprise` |
| npm package | `@dinglebear/rapprise` |
| npm binary aliases | `apprise-rmcp`, `rapprise` |
| MCP tool | `apprise` |
| Config home | `~/.apprise` on hosts, `/data` in containers |
| Env prefixes | `APPRISE_*`, `APPRISE_MCP_*`, `APPRISE_RMCP_*` for npm launcher controls |
The repo and npm package use the RMCP family name, while the shipped binary uses
the short Rust CLI name `rapprise`.
## Capabilities And Boundaries
- Send notifications through tags configured in the upstream Apprise API server.
- Send one-off notifications to Apprise URL schemas with `notify_url`.
- Check upstream Apprise API health.
- Expose the `send_alert` prompt for critical alert workflows.
- Provide setup and doctor commands for local plugin/runtime checks.
| This repo owns | Apprise owns | Explicitly out of scope |
|---|---|---|
| MCP/CLI projection, request validation, auth policy, response shaping, setup checks, and transport wiring. | Notification delivery, configured destinations, tags, delivery backend credentials, upstream API semantics. | Destination storage, scheduling, retry policy beyond upstream behavior, multi-tenant sandboxing, arbitrary webhook relay behavior. |
## Install
| Path | Command | Best for | Notes |
|---|---|---|---|
| npm / npx | `npx -y @dinglebear/rapprise --help` | Linux/Windows x86_64 clients. | Verifies the release archive SHA-256 before atomic install. |
| Release installer | [Verified installer procedure](#verified-release-installer) | Linux x86_64 without Node. | Verifies checksum and offline provenance before executing installer code. |
| Docker / Compose | `docker compose up -d` | Shared HTTP MCP deployments. | Reads `.env` and exposes container port `40050`. Needs the external network first — see [Deployment](#docker--compose). |
| Build from source | `cargo build --release` | Development and audits. | Produces `target/release/rapprise`. |
| Plugin | `just build-plugin && claude plugin install ./plugins/apprise` | Claude Code from this checkout. | Bundled `rapprise` stdio plugin. Ships no hooks. |
Releases publish SHA-256 files and offline GitHub build-provenance bundles. The
installers verify both the checksum and provenance identity with GitHub CLI 2.68+.
The npm launcher supports x86_64 Linux and Windows release assets.
### npm / npx
Run the stdio MCP server or CLI without a manual binary install:
```bash
npx -y @dinglebear/rapprise --help
npx -y @dinglebear/rapprise mcp
npx -y @dinglebear/rapprise health --json
```
The npm package downloads `rapprise` during `postinstall`. Override download
behavior only when testing packaging:
| Variable | Purpose |
|---|---|
| `APPRISE_RMCP_SKIP_DOWNLOAD=1` | Skip postinstall binary download. |
| `APPRISE_RMCP_VERSION` or `APPRISE_RMCP_BINARY_VERSION` | Select the GitHub Release tag. |
| `APPRISE_RMCP_REPO` | Select the GitHub repo used for release downloads. |
| `APPRISE_RMCP_RELEASE_BASE_URL` | Select a custom release base URL. |
Verified binary installation is fail-closed and pins the repository, release
workflow, and source tag. Install `gh` 2.68 or newer before using npm or the
release installer.
### Verified Release Installer
Download the installer and its verification material without executing it,
then verify both integrity and provenance before running it:
Replace `version` with the release tag you want; `v0.2.0` is current.
```bash
version=v0.2.0
base="https://github.com/dinglebear-ai/rapprise/releases/download/${version}/rapprise-installer.sh"
curl -fsSLO "$base"
curl -fsSLO "$base.sha256"
curl -fsSLO "$base.sigstore.json"
sha256sum --check rapprise-installer.sh.sha256
gh attestation verify rapprise-installer.sh \
--repo dinglebear-ai/rapprise \
--bundle rapprise-installer.sh.sigstore.json \
--signer-workflow dinglebear-ai/rapprise/.github/workflows/release.yml \
--source-ref "refs/tags/${version}" \
--deny-self-hosted-runners
APPRISE_RMCP_VERSION="$version" bash rapprise-installer.sh
```
The `--repo` and `--signer-workflow` values must match the identity recorded in
the attestation at build time. Releases cut before the repository moved to the
`dinglebear-ai` org are attested as `jmagar/rapprise`; use `dinglebear-ai/rapprise`
for releases cut afterwards.
The npm launcher supports Windows x86_64 only when GitHub CLI 2.68+ is installed
and `gh.exe` is available on `PATH`; provenance verification is not skipped.
### Build From Source
```bash
git clone https://github.com/dinglebear-ai/rapprise
cd rapprise
cargo build --release
./target/release/rapprise --help
```
Minimum supported Rust version: 1.90. Rust edition 2021. The Cargo workspace has
two members: the root `apprise-mcp` package and `xtask`.
## Quickstart
### 1. Start Or Point At Apprise API
The default upstream is `http://localhost:8000`.
```bash
docker run --rm -p 8000:8000 caronc/apprise:latest
```
Use `APPRISE_URL` when the API server is elsewhere:
```bash
export APPRISE_URL=http://100.120.242.29:8766
```
Set `APPRISE_TOKEN` only when your Apprise API server requires bearer auth:
```bash
export APPRISE_TOKEN=...
```
### 2. Run A Safe CLI Call
```bash
npx -y @dinglebear/rapprise health --json
```
### 3. Start Loopback HTTP MCP
```bash
APPRISE_MCP_HOST=127.0.0.1 npx -y @dinglebear/rapprise serve
```
In another shell:
```bash
curl -sf http://127.0.0.1:40050/health
```
### 4. Make A First MCP Call
```bash
curl -s -X POST http://127.0.0.1:40050/mcp \
-H "Content-Type: application/json" \
-H "Accept: application/json, text/event-stream" \
-d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"apprise","arguments":{"action":"health"}}}'
```
## Client Configuration
### Claude Code Stdio
```json
{
"mcpServers": {
"apprise": {
"command": "npx",
"args": ["-y", "apprise-rmcp", "mcp"],
"env": {
"APPRISE_URL": "http://localhost:8000"
}
}
}
}
```
### Claude Code HTTP
```json
{
"mcpServers": {
"apprise": {
"type": "http",
"url": "http://127.0.0.1:40050/mcp",
"headers": {
"Authorization": "Bearer ${APPRISE_MCP_TOKEN}"
}
}
}
}
```
### Codex / Labby Gateway
Register Apprise through Labby as an HTTP upstream when sharing one long-running
server, or run it directly as stdio for local-only use.
```toml
[mcp_servers.apprise]
command = "npx"
args = ["-y", "apprise-rmcp", "mcp"]
```
### Generic MCP JSON
```json
{
"command": "rapprise",
"args": ["mcp"],
"env": {
"APPRISE_URL": "http://localhost:8000"
}
}
```
Do not put `APPRISE_TOKEN`, OAuth secrets, SSH keys, passwords, or upstream
bearer tokens in MCP tool arguments. Use env, config files, or the MCP client's
secret storage.
## Runtime Surfaces
| Surface | Status | Entry point | Purpose |
|---|---:|---|---|
| MCP stdio | Supported | `rapprise mcp`, `npx -y @dinglebear/rapprise mcp` | Local child-process MCP clients. |
| MCP HTTP | Supported | `rapprise serve`, `POST /mcp` | Streamable HTTP MCP for local or shared server deployments. |
| Liveness | Supported | `GET /health` | Always unauthenticated. |
| Readiness / status | Supported | `GET /ready`, `GET /status` | Behind the same auth layer as `/mcp`. |
| OAuth discovery | Conditional | `/mcp/.well-known/*` and `lab-auth` routes | Mounted only when `auth_mode = oauth`. |
| CLI | Supported | `rapprise <command>` | Scriptable parity and debugging. |
| Prompt | Supported | `send_alert` | Reusable critical-alert workflow. |
| REST API | Not shipped | N/A | Apprise API already owns the REST API. |
| Web UI | Not shipped | N/A | Apprise API already owns the web UI. |
## MCP Tool Reference
One MCP tool is exposed: `apprise`. Pass the required `action` argument to select
the operation.
### Read Actions
| Action | Description | Required params | Optional params |
|---|---|---|---|
| `health` | Check Apprise API server health. | none | none |
| `status` | Return authenticated deployment and runtime status. | none | none |
| `help` | Return built-in markdown tool help. | none | none |
### Write Actions
| Action | Description | Required params | Optional params |
|---|---|---|---|
What people ask about rapprise
What is dinglebear-ai/rapprise?
+
dinglebear-ai/rapprise is mcp servers for the Claude AI ecosystem. Rust MCP server and CLI for Apprise notification fan-out across dozens of delivery backends, with stdio/HTTP transports, auth, and plugin packaging. It has 1 GitHub stars and was last updated today.
How do I install rapprise?
+
You can install rapprise by cloning the repository (https://github.com/dinglebear-ai/rapprise) or following the README instructions on GitHub. ClaudeWave also provides quick install blocks on this page.
Is dinglebear-ai/rapprise safe to use?
+
dinglebear-ai/rapprise has not been audited yet by our security agent. Review the original repository on GitHub before using it in production.
Who maintains dinglebear-ai/rapprise?
+
dinglebear-ai/rapprise is maintained by dinglebear-ai. The last recorded GitHub activity is from today, with 1 open issues.
Are there alternatives to rapprise?
+
Yes. On ClaudeWave you can browse similar mcp servers at /categories/mcp, sorted by popularity or recent activity.
Deploy rapprise to your cloud
Ship this repo to production in minutes. Each platform spins up its own environment with editable env vars.
Maintain this repo? Add a badge to your README
Drop the badge into your GitHub README to show it's tracked on ClaudeWave. Each badge links back to this page and reflects the live Trust Score.
[](https://claudewave.com/repo/dinglebear-ai-rapprise)<a href="https://claudewave.com/repo/dinglebear-ai-rapprise"><img src="https://claudewave.com/api/badge/dinglebear-ai-rapprise" alt="Featured on ClaudeWave: dinglebear-ai/rapprise" width="320" height="64" /></a>More MCP Servers
Fair-code workflow automation platform with native AI capabilities. Combine visual building with custom code, self-host or cloud, 400+ integrations.
User-friendly AI Interface (Supports Ollama, OpenAI API, ...)
An open-source AI agent that brings the power of Gemini directly into your terminal.
The fastest path to AI-powered full stack observability, even for lean teams.
Real-time global intelligence dashboard. AI-powered news aggregation, geopolitical monitoring, and infrastructure tracking in a unified situational awareness interface
🕷️ An adaptive Web Scraping framework that handles everything from a single request to a full-scale crawl!