pacer-mcp
This MCP server provides tools to interact with the Pacer API for portfolio management, performance metrics, and PMS integration, but it is now retired and users should switch to the in-process Pacer MCP connector.
README
pacer-mcp
🪦 RETIRED. This standalone PAT-over-stdio server no longer reaches the Pacer API. Every tool returns a notice to switch to the in-process ("inline") Pacer MCP connector built into core, offered as the hosted "Pacer" connector on claude.ai / Claude Cowork (tools appear as
mcp__…Pacer__…). The standalone binary called the API over the network atportal.pacerrev.io, which now redirects unauthenticated API calls to an HTML login page; the in-process connector dispatches to the same/api/v1handlers with no network hop and can't drift from the API host. Remove this server from your MCP client config. The setup and tool docs below are kept for historical reference only.
An MCP server that exposes pacer/core API endpoints as native tools for Claude Code (and any other MCP-aware client). One binary, stdio transport, no fuss.
Install
macOS (Homebrew)
brew install STR-Consulting/tap/pacer-mcp
Windows (Scoop)
scoop bucket add pacer https://github.com/STR-Consulting/scoop-bucket
scoop install pacer-mcp
Non-programmer? See docs/windows-setup.md for a copy-paste prompt to hand to your AI assistant.
Linux (binary tarball)
For server/automation use. Replace <version> and pick amd64 or arm64:
curl -fsSL https://github.com/STR-Consulting/mcp/releases/latest/download/pacer-mcp_<version>_linux_amd64.tar.gz | tar -xz
sudo install pacer-mcp /usr/local/bin/
Or use the OCI image: ghcr.io/str-consulting/pacer-mcp:latest (linux/amd64 + linux/arm64).
From source
go install github.com/STR-Consulting/pacer-mcp@latest
Setup
Add to your MCP config (e.g. .mcp.json in your project or your Claude Code user config):
{
"mcpServers": {
"pacer": {
"command": "pacer-mcp",
"env": {
"PACER_CORE_URL": "https://portal.pacerrev.io",
"PACER_CORE_TOKEN": "pat_..."
}
}
}
}
| Env var | Description | Default |
|---|---|---|
PACER_CORE_URL |
Base URL of the pacer/core app (PAT JSON API is mounted here) | https://portal.pacerrev.io |
PACER_CORE_TOKEN |
Personal access token, format pat_... |
(unset) |
Run health_check after install to confirm the server can reach core.
PATs are minted by a core admin via pacer pat create --user <email> --label <name>; they require an employee-or-higher role and are sent as Authorization: Bearer pat_....
Tools
Each tool's full description (caveats, when to use, args) is returned to the MCP client at registration time — agents see them at tool-selection time. The list below is a quick index.
Operational
| Tool | What it does |
|---|---|
health_check |
Pings the Pacer API and reports config + reachability |
Portfolio fundamentals
| Tool | What it does |
|---|---|
list_briefable_portfolios |
Enumerate active portfolios (optional q name filter) |
list_portfolio_teams |
Bulk: notification team (RM/RD/Jon) for every portfolio |
get_portfolio_team |
Notification team for a single portfolio |
list_portfolio_units |
Unit roster: bedrooms, type, managed/active, location |
list_portfolio_reservations |
Reservations in a date range (by check_in/check_out/booked_on) |
Performance
| Tool | What it does |
|---|---|
get_portfolio_pacing |
Recent reservations w/ YoY rent/ADR/ABW/LOS + anomaly score |
get_portfolio_metrics_ytd |
CY vs PY YTD: revenue, ADR, occupancy, RevPAR, LOS, count |
get_portfolio_market_metrics |
One-month CY vs PY + market benchmark deltas; optional decomposition |
Guesty PMS
| Tool | What it does |
|---|---|
guesty_pricing_config |
Per-unit pricing intent: base price, fees, min/max nights, factors, channel settings |
guesty_reservation_promotions |
Channel-applied promos on reservations in a month (Airbnb, Vrbo, etc.) |
Client health
| Tool | What it does |
|---|---|
get_client_health_brief |
Latest (or dated) sentiment brief for a portfolio |
upsert_client_health_brief |
Log a lightweight sentiment brief (1-5 + stage + payload) |
list_client_health_briefs |
Dashboard view: latest brief per portfolio as of a date |
get_client_health_scoring_config |
Scoring weights, labels, and tier thresholds |
upsert_intel_brief |
Publish a full intel brief (Postgres + ClickUp task + BigQuery mirror) |
KeyData
| Tool | What it does |
|---|---|
list_managed_keydata_units |
Pacer-managed unit UUIDs for a KeyData customer account |
Not wrapped:
POST /portfolios/{p}/intel-brief/attachments(multipart binary upload). MCP has no natural multipart story.
Development
go build -o pacer-mcp .
go test ./...
golangci-lint run --fix ./...
Releases are automated — push a v* tag and GitHub Actions builds darwin-arm64, windows-amd64, and linux-{amd64,arm64} binaries (plus a ghcr.io/str-consulting/pacer-mcp OCI image), then updates the Homebrew tap and Scoop bucket and publishes to the MCP Registry.
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。