Context7
Up-to-date docs for AI agents
제작자가 쓴 설명(영어 원문)
Send tasks and notifications to people's phones. Answers build a queryable knowledge base.
이 세 가지는 절대 하나의 점수로 합치지 않아요. 의견, 측정, 인기는 서로 다른 것이니까요.
이 서버는 아직 쉬운 말로 된 요약을 쓰지 못했어요. 아래는 레지스트리에서 읽어 온 사실이에요.
쓰고 있는 AI 앱을 고른 다음 설정을 복사하세요.
SP_API_TOKEN, SP_INTEGRATION_TOKEN, SP_KEYS, SP_BASE_URL, SP_MAX_WAIT_SECONDS비밀로 표시된 값은 내 계정의 열쇠예요. 절대 다른 사람과 공유하거나 채팅에 붙여 넣지 마세요.
필요한 것
어떤 앱을 쓰세요?
위에서 앱을 고르세요. 기억해 두었다가 다음부터는 버튼에 “…에 설치”라고 표시해요.
웹, 데스크톱, 모바일의 Claude에서 쓸 수 있어요:
https://mcp.simplepu.sh/mcpClaude Desktop → Settings → Developer → Edit Config를 열고 claude_desktop_config.json에 이 내용을 붙여 넣으세요
{
"mcpServers": {
"simplepu": {
"command": "npx",
"args": [
"-y",
"@simplepush/mcp"
],
"env": {
"SP_API_TOKEN": "your-value",
"SP_INTEGRATION_TOKEN": "your-value",
"SP_KEYS": "your-value",
"SP_BASE_URL": "your-value",
"SP_MAX_WAIT_SECONDS": "your-value"
}
}
}
}터미널에서 이 명령어를 실행하세요
claude mcp add --transport http simplepu https://mcp.simplepu.sh/mcpclaude mcp add simplepu --env SP_API_TOKEN=your-value --env SP_INTEGRATION_TOKEN=your-value --env SP_KEYS=your-value --env SP_BASE_URL=your-value --env SP_MAX_WAIT_SECONDS=your-value -- npx -y @simplepush/mcp실행하기 전에 빈 값을 채워 주세요.
프로젝트의 .cursor/mcp.json 또는 ~/.cursor/mcp.json에 붙여 넣으세요
{
"mcpServers": {
"simplepu": {
"url": "https://mcp.simplepu.sh/mcp"
}
}
}프로젝트의 .cursor/mcp.json 또는 ~/.cursor/mcp.json에 붙여 넣으세요
{
"mcpServers": {
"simplepu": {
"command": "npx",
"args": [
"-y",
"@simplepush/mcp"
],
"env": {
"SP_API_TOKEN": "your-value",
"SP_INTEGRATION_TOKEN": "your-value",
"SP_KEYS": "your-value",
"SP_BASE_URL": "your-value",
"SP_MAX_WAIT_SECONDS": "your-value"
}
}
}
}브라우저가 앱을 열지 물어본 다음, 앱이 설정을 보여 주고 확인을 요청해요. 키 값은 비워 두니 직접 입력하세요.
프로젝트의 .vscode/mcp.json에 붙여 넣으세요
{
"servers": {
"simplepu": {
"type": "http",
"url": "https://mcp.simplepu.sh/mcp"
}
}
}프로젝트의 .vscode/mcp.json에 붙여 넣으세요
{
"servers": {
"simplepu": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@simplepush/mcp"
],
"env": {
"SP_API_TOKEN": "your-value",
"SP_INTEGRATION_TOKEN": "your-value",
"SP_KEYS": "your-value",
"SP_BASE_URL": "your-value",
"SP_MAX_WAIT_SECONDS": "your-value"
}
}
}
}브라우저가 앱을 열지 물어본 다음, 앱이 설정을 보여 주고 확인을 요청해요. 키 값은 비워 두니 직접 입력하세요.
ChatGPT는 인터넷에 있는 서버에만 연결할 수 있어요:
https://mcp.simplepu.sh/mcp메뉴 이름은 요금제나 시기에 따라 조금 다를 수 있어요.
선택한 앱은 이 기기의 쿠키에만 저장돼요. 계정은 필요 없어요.
SP_BASE_URL envSP_MAX_WAIT_SECONDS env이 설정은 제작자의 레지스트리 정보로 만들었어요. README와 다르면 README를 따르세요.
서버가 도구 목록을 보여 주기 전에 로그인을 요구해요.
누구나 볼 수 있는 정보를 점검했고, 코드는 읽지 않았어요. 해당하지 않는 항목은 개수에서 뺐어요. · 2026년 11월 6일 점검
레지스트리 이름이 도메인 simplepu.sh 기준으로 확인되었어요(레지스트리가 도메인 소유를 확인해요)
아직 GitHub에서 이 저장소를 읽어 오지 않았어요
아직 GitHub에서 이 저장소를 읽어 오지 않았어요
208ms 만에 응답(HTTP 401)
OAuth 로그인을 지원한다고 밝혀서, AI 앱이 로그인 화면으로 안내해 줘요
아직 점검하지 않았어요. 엔드포인트를 점검할 때 도구 목록을 읽어요
레지스트리에서 이 카테고리의 활성 서버 전체와 비교했어요. 스타는 게시자가 저장소를 소유한 경우에만 세요.
해당하는 기본 점검 3개 중 3개를 통과했어요. 코드는 읽지 않았어요. 키나 토큰을 넣기 전에 서버가 어떤 권한을 요구하는지 확인하고, 꼭 필요한 권한만 있는 키를 쓰세요.
입력해야 할 값: SP_API_TOKEN, SP_INTEGRATION_TOKEN, SP_KEYS
둘 다 가능해요. 내 컴퓨터에 설치해도 되고, 클라우드 버전에 연결해도 돼요.
GitHub 저장소는 마지막으로 – 변경됐고, 최신 버전(0.5.0)은 그저께 레지스트리에 올라왔어요.
Claude Desktop, Claude Code, Cursor, VS Code처럼 MCP를 지원하는 앱에서 쓸 수 있어요. URL로 연결되기 때문에, 원격 MCP 서버를 지원하는 앱(예: 개발자 모드의 ChatGPT)에서도 쓸 수 있어요.
같은 카테고리에는: Context7, Codebase Memory, Exa
같은 카테고리에서 다른 제작자가 만든 서버예요. 스타 순으로 정렬했어요.
카드 보는 법:★ GitHub 스타i세 기둥: 리뷰 · 점검 · 사용 정도i업데이트 = 최신 버전을 레지스트리에 공개한 날i
Up-to-date docs for AI agents
Codebase knowledge graph for AI agents — 162 languages, sub-ms queries, 99% fewer tokens.
Connect AI agents to Exa for web search, content fetching, and multi-step research.
제작자의 저장소에서 쓰인 그대로 가져왔어요. PopNx MCP가 검토하거나 보증한 내용이 아니에요. HTML, 이미지, 스크립트는 제거했어요. 링크를 누르면 이 사이트를 벗어나요.
MCP server for Simplepush. An agent sends tasks and notifications to people's phones and waits for the answers, appends follow-ups, cancels mistakes, and queries the record of everything that came back. Tasks, answers, replies, declines, and ad-hoc submissions stay queryable, so the stream doubles as a knowledge base of the work: what was asked, who answered what, and what is still open.
| Tool | What it does |
|---|---|
send_notification | Push for alerts and heads-ups. Can carry one input (a text input, a choice, or action buttons); returns right after sending unless wait_seconds is set. A copy per recipient by default — the result lists one notification_id per person; shared: true sends one all recipients see. |
send_task | Sends a task and waits for the answers. Takes the full input set (text, choice, actions, slider, photo, voice recording, file, location), a tag, a reply mode, expires_at, Markdown and links. A copy per recipient by default: waits for everyone, reports per-recipient results with the rest pending at the deadline. shared: true sends one task all recipients answer together, first answer wins. Every result carries the append_token. |
append_subtask | Appends a follow-up question or checklist item to a task or group, using the append_token from send_task. Never waits. |
cancel_task | Withdraws a task (tsk_), subtask (sub_) or whole group (grptsk_) by id, with an optional reason and note. |
get_task_answer | Resolves a task_id from an earlier send_task, or a subtask_id from append_subtask. |
get_notification_answer | Resolves what came back on a notification_id from an earlier send_notification with an input. |
query_tasks | One page of tasks as summaries: tag, topic, status, recipients, input kinds, attachment kinds, reply mode, sent time, subtask counts. Filters: status, time window, topic, member, group. |
get_task | One task with all its subtasks: questions, answers, replies, declines, file ids. |
get_group_status | Per-recipient status of a task group (grptsk_): who has answered, who has not. |
query_events | The org's activity history: answers, replies, submissions, declines, cancellations, expiries, with who and when. Last 7 days by default. |
query_submissions | Ad-hoc reports from the field (text, photo, file, voice, location), with who and when. Last 7 days by default. |
search_knowledge | Full-text and location search over everything: tasks and notifications, what people answered and replied, and ad-hoc submissions, across all time. Literal word matching, plus stemming in up to three languages; or a coordinate plus radius (or a polygon of corner points), returning the locations recorded within it, nearest first. Each hit carries the id to read in full (a group send hits once, as its grptsk_ id) and a snippet or a point with its distance. Plaintext records only. |
get_activity | Everything going on for one member or the whole org in one call: open/declined/expired tasks plus the window's answers, replies, declines, cancellations and submissions. The tool for "any problems with X?" and "catch me up". |
download_attachment | Fetches a photo, voice recording or file by scope_id (the tsk_/sub_/sbm_ id it belongs to) and file_id (an answer's inputId, a reply or submission file's id). Images are returned inline. |
The query tools work in both modes: an organization credential (an integration token, or the hosted connector signed in as an organization admin) reads the whole organization, a personal one (an API token, or the hosted connector with a personal API token) reads your personal stream (subscription required on personal accounts). On the hosted transport they need the read scope; download_attachment additionally needs files:read. Encrypted content is decrypted with the keys the server holds; anything it cannot open is left as ciphertext and reported.
A timeout in send_task is not a cancellation. The questions stay live on the phones, and answers are available later via get_task_answer or get_group_status. inputs says how the task is answered: text, single or multi choice, action buttons, slider, photo, voice recording, file, location. send_task waits 90 seconds by default; per-call wait_seconds raises that, up to SP_MAX_WAIT_SECONDS.
The server runs in one of two modes, decided by which credential you set.
Get an API token from the API Token screen in the Simplepush app, then:
{
"mcpServers": { "simplepush": { "command": "npx", "args": ["-y", "@simplepush/mcp"], "env": { "SP_API_TOKEN": "your-token" } } } }
Or with Claude Code:
claude mcp add simplepush --env SP_API_TOKEN=your-token -- npx -y @simplepush/mcp
Messages go to your own devices. Pass a topic to deliver to that topic's holders instead, as a single shared task where the first answer wins.
Create an integration token with the CLI (admin vault unlocked):
sp integration create
claude mcp add simplepush --env SP_INTEGRATION_TOKEN=spi_... -- npx -y @simplepush/mcp
Org mode unlocks full targeting: every send takes exactly one of topic, member (name or usr_ id), or broadcast. Only the tools the token's scopes cover are listed: a token minted with read files:read exposes the query and download tools and none of the send tools, which keeps the tool schemas a read-only agent carries per request to a minimum. If the org has encryption enabled, the token's second half unwraps the org master keys at startup and all sends are end-to-end encrypted; answers are decrypted the same way.
If both credentials are set, the integration token wins and a note goes to stderr. MCP clients spawn stdio servers with your shell environment attached, so an exported SP_API_TOKEN from unrelated CLI work rides along; setting SP_INTEGRATION_TOKEN is always deliberate.
| Variable | Default | Meaning |
|---|---|---|
SP_API_TOKEN | one credential required | Personal API token from the app. |
SP_INTEGRATION_TOKEN | one credential required | Org integration token from sp integration create. |
SP_KEYS | unset | Personal encryption keys, see below. |
SP_BASE_URL | https://api.simplepu.sh | API origin. Point at http://localhost:8000 for local development. |
SP_MAX_WAIT_SECONDS | 900 | Ceiling on how long send_task may block. Per-call wait_seconds is clamped to it. |
SP_POLL_INTERVAL_MS | 2000 | Gap between answer polls while blocking. |
SP_SCOPES | unset | Space- or comma-separated scope codes (send, read, files:read); only the tools they cover are listed. Narrows an integration token's own scopes; the only way to trim the listing for a personal API token, which the backend always treats as full access. |
Org mode encrypts automatically when the org has encryption enabled, as described above.
Personal sends are plaintext unless you provide keys. Encrypted and plaintext tasks are distinguishable at a glance in the app: encrypted ones carry the lock icon.
Keys are exported from the Simplepush app: Copy Key for Integrations on the Encryption screen for the Personal Password key, and the copy action next to a topic on the Topics screen for topic keys.
SP_KEYS is a comma-separated list. A bare base64 key is the Personal Password key, used for self-sends. topic=key binds a key to one topic:
SP_KEYS="AbC...=" # default key only
SP_KEYS="alerts=AbC...=,deploys=XyZ...=" # two topic keys SP_KEYS="AbC...=,alerts=XyZ...=" # both
With a matching key, sends to that target are encrypted and answers are decrypted. An answer the server holds no key for comes back with undecryptable: true rather than silently wrong.
Keys only. There is no SP_PASSWORD.
simplepush-mcp-http serves the same tools over Streamable HTTP as an OAuth resource server. No ambient credential: each request carries its own access token, verified by introspection against the authorization server and audience-checked against SP_CANONICAL_URI. Discovery metadata is served at /.well-known/oauth-protected-resource; the MCP endpoint is /mcp. GET /status is a public summary for uptime monitoring: 200 when the authorization server and the API both answer, 503 otherwise, with the same shape as the backend's /v1/status. Tools are scope-gated: send for send_notification, send_task, append_subtask and cancel_task, read for the answer and query tools, files:read for download_attachment. A grant lists only the tools its scopes cover; a direct call for anything else is refused with a challenge naming the missing scope.
Hosted mode is personal OAuth grants only. Integration tokens are not accepted there by design, and hosted sends are plaintext: a server that could decrypt for you would not be end-to-end.
| Variable | Default | Meaning |
|---|---|---|
SP_INTROSPECTION_SECRET | required | Bearer for the AS introspection endpoint. Must match the backend's oauth.introspectionSecret. |
SP_CANONICAL_URI | required | Public URL clients reach this server at, e.g. https://mcp.simplepu.sh/mcp. Tokens are audience-bound to it. |
SP_OAUTH_ISSUER | https://api.simplepu.sh | Origin of the authorization server, as clients see it. |
SP_AUTH_SERVER_URL | SP_OAUTH_ISSUER | Where this server reaches the authorization server for token introspection. Set it when the backend is closer on an internal address. |
SP_BASE_URL | https://api.simplepu.sh | API origin this server calls. |
SP_MCP_PORT | 8787 | Listen port. |
npm install
npm run build npm run typecheck
Smoke-test against a local backend without an MCP client:
npx @modelcontextprotocol/inspector --cli node dist/index.mjs \
-e SP_API_TOKEN=testtoken5678 -e SP_BASE_URL=http://localhost:8000 \ --method tools/list
Note that the Inspector does not forward your shell environment to the spawned server. Pass credentials with -e, not by exporting them.