arcai-hr
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@arcai-hrshow me candidates for the software engineer opening"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
arcai-hr
A reusable API/action foundation for ArcAI vertical applications, proved out with HR as the first vertical.
Architecture
UI Layer (React + Tailwind, GrowthArc tokens)
| calls
MCP Layer -- FastMCP / ARC MCP
| exports the same catalog as
HR Vertical Layer -- services + ARC Actions
| built on
Base/API Layer -- reusable FastAPI coreBase layer (
app/core,app/db,app/actions,app/mcp): config, DB plumbing, identity/ActionContext, error mapping, and a generic route generator that mounts one FastAPI endpoint perarcai_core.ArcActionin anyActionCatalog-- vertical-agnostic.HR vertical layer (
app/verticals/hr, added in Phase 2): models, services, and realArcActions built onarcai_core.MCP layer: the same catalog exported as FastMCP tools via the generic exporter in
app/mcp.UI layer: a minimal React + TypeScript + Tailwind app (Phase 4), following
docs/GrowthArc-UIUX-Design-System-Guidelines.md.
arcai-runtime (auth, approvals, SSE, audit) is explicitly out of scope for
this round -- the pieces above are small, dependency-free implementations of
the same pattern, built from scratch to learn it.
Related MCP server: Homerun ATS MCP Server
Setup
This project depends on arcai-core as an editable sibling checkout. Clone
it next to this repo:
Projects/
arcai-core/
arcai-hr/Then, from this directory:
uv syncRun
uv run uvicorn main:app --reloadmain.py creates the HR tables (SQLite, zero setup) and mounts the HR action
catalog. Interactive docs at http://localhost:8000/docs; the catalog
manifest at GET /actions.
Demo data (4 openings, 9 candidates across every stage, interviews, feedback):
uv run python seed.pySafe to re-run; delete arcai_hr.db to start from scratch.
MCP
The same catalog, exported as MCP tools over stdio (one tool per action,
flat arguments from each action's input model, read_only -> readOnlyHint):
uv run python mcp_server.pyUI
React + TypeScript + Tailwind v4 (+ Radix Dialog, Lucide), styled from the
GrowthArc tokens in ui/src/index.css. Talks to the FastAPI layer at
VITE_API_URL (default http://localhost:8000).
cd ui
npm install
npm run dev # http://localhost:5173 -- run the API alongside itTest
uv run pytestThis server cannot be deployed
Maintenance
Related MCP Connectors
Human-input bridge for AI agents with voice-first answer links, MCP tools, and HTTP APIs.
Official 100Hires MCP: AI ATS & Recruitment Software for candidates, jobs, applications, interviews.
HR and recruiting: score a job description, or generate a structured role profile. No API key.
- mcpOAuthcom.curviate
LinkedIn actions for AI agents: search, messaging, posts and invites, as hosted MCP tools.
Related MCP Servers
AlicenseBqualityFmaintenanceOfficial Model Context Protocol server for 100Hires — the applicant tracking system for recruiting teams. Exposes the full 100Hires API v2 as 130 MCP tools, enabling AI assistants to manage candidates, jobs, applications, interviews, messages, and more.10016 npm1MIT- FlicenseNot gradedqualityDmaintenanceMCP server that connects AI agents to Homerun, an applicant tracking system. Enables AI-assisted candidate review: browse vacancies, inspect applications, and leave review notes.-
- AlicenseNot gradedqualityDmaintenanceExposes core recruiting tools such as candidate ranking, profile retrieval, honeypot audits, and job description parsing via stdio protocol.MIT
- AlicenseAqualityCmaintenanceLets any MCP client drive the recruiting workflow in natural language: create roles, screen CVs, schedule and manage candidate interviews, and read back scored reports with transcripts.18MIT