Skip to main content

Announcing StackOne Defender: leading open-source prompt injection guard for your agent Read More

Reference Check Automation

Reference Checks, Automated

Use StackOne to connect your AI agent to your ATS, email, and form tools to automate reference check requests, reminders, and summaries.

ClaudeOpenAILangChainVercel

AI Agents

Connect

MCP and A2A to REST, SOAP, and proprietary APIs.

Optimize

Tool discovery, data shaping, and reliable execution.

Secure

Scoped permissions, audit trails, and observability.

StackOne Integration Layer

get_candidate
greenhouse greenhouse
get_application
greenhouse greenhouse
send_email
gmail gmail
update_candidate
greenhouse greenhouse
send_message
slack slack

What Can AI Agents Do for Reference Check Automation?

Your agent detects the stage change, pulls referee contacts, sends personalized forms, collects responses, summarizes feedback, and updates the candidate profile.

01

Detect Stage Change

Listen for the candidate moving to the "Reference Check" stage in Greenhouse, Lever, or Workday via webhooks.

Greenhouse
02

Pull Referee Contacts

Retrieve referee names, emails, and relationships from the candidate's ATS profile. Works across Greenhouse, Lever, Workday, and iCIMS.

Lever
03

Send Reference Forms

Send personalized reference check forms via SurveyMonkey or Typeform to all referees over Gmail or Slack. Automate reminders for non-respondents.

SurveyMonkey
04

Synthesize Feedback

Parse completed reference responses and generate a summary highlighting strengths, concerns, and patterns across all referees for quick recruiter review.

05

Update Candidate Profile

Attach the summary report and individual responses to the candidate's ATS profile. Notify the recruiter via Slack or Microsoft Teams.

Slack
06

Advance Candidate Stage

Optionally update the candidate's ATS stage to "Reference Check Complete" and log the entire workflow for audit purposes.

Why Building a Good Reference Check Agent Is Hard

Connecting to Many ATS and Outreach Systems

Building a reference check agent across Greenhouse, Lever, Workday, iCIMS, and SmartRecruiters means maintaining separate integrations — each with unique auth, pagination, and rate limits. Add email and form tools and the connector count doubles.

Token Cost Grows with Tool Definitions

Without search-first tool selection, the agent pre-loads every ATS and messaging action definition into context, burning tokens and money — worse when supporting many customers with different tool stacks.

Unreliable Stage-Change Events Across ATS Providers

Each ATS delivers webhooks differently — some poll, some push, some delay. Building reliable triggers for the reference check stage requires custom event infrastructure per provider.

Untrusted Referee Inputs Risk Prompt Injection

Reference responses are external, uncontrolled text fed directly into the agent. Without input guards, malicious or malformed responses can hijack agent behavior — skipping checks or leaking candidate data.

How StackOne Makes Reference Check Agents Possible

Everything your reference check agent needs to connect to ATS platforms, send forms, collect feedback, and update candidate profiles — with the controls IT demands.

200+ connectors with 10K+ agent-optimized actions

Pre-built ATS connectors for Greenhouse, Lever, Workday, iCIMS, and more — each with full native actions mapped with agent instructions. Plus connectors for Gmail, SurveyMonkey, Typeform, and Slack.

Managed Auth handles credentials across providers

Managed Auth handles credentials across providers

OAuth flows, API keys, and token refresh managed per tenant for every connected ATS, form tool, and messaging system — agents never touch raw credentials.

Search and execute finds the right action

Agent searches StackOne's action catalog by natural language and executes the matching ATS or messaging action — no pre-loading thousands of tool definitions.

Managed Webhooks deliver ATS stage events consistently

StackOne subscribes to candidate stage-change events across connected ATS platforms, handling registration, retries, and delivery differences so the agent receives a consistent event stream.

Connector Studio extends to any system

Connector Studio extends to any system

Build custom connectors for unsupported ATS platforms or niche survey tools via REST, SOAP, or GraphQL — no waiting on vendor support.

Defender blocks prompt injection from referee responses

StackOne Defender screens inbound referee text for injection attempts before the agent processes it, preventing adversarial content from manipulating reference check behavior.

You Control What the Agent Can Do

You Control What the Agent Can Do

Scoped permissions define exactly which candidate fields the agent reads and which ATS actions it can trigger. Full audit trail of every operation.

Connect Any Agent to Automate Reference Checks

claudeopenailangchainvercelcrewaipydantic

Any Agent Framework

Claude, OpenAI, LangChain, Vercel AI SDK, CrewAI, Pydantic AI — StackOne works with every major agent framework out of the box.

flowisen8nmakesanamicrosoft-copilot

Any Agent Builder

Whether you're building with code, a visual builder, or an enterprise platform — StackOne provides the integration layer your agent needs.

Any Protocol

Pick the protocol that fits your stack. Tool calling, direct API integration, agent-to-agent messaging, or structured action workflows — all supported out of the box.

Connect Your Agent to Your Recruiting Stack

Start building in minutes. MCP connectors to every system your agent needs.

Frequently Asked Questions

The hardest parts are connecting to each customer's ATS (Greenhouse, Lever, Workday all use different APIs and auth), reliably detecting stage changes to trigger outreach, and safely processing untrusted referee responses without exposing the agent to manipulation. Most teams underestimate the integration overhead.
Three stand out: each ATS (Greenhouse, Lever, Workday) uses different APIs, auth, and webhook formats requiring per-platform connectors. Enriching with form tool data adds more integrations. And referee text from external respondents creates prompt injection risk that can manipulate agent behavior.
You need an integration layer with pre-built connectors for each ATS, where each provider's native actions are mapped with agent-ready instructions. StackOne offers 200+ connectors covering major ATS platforms, so the agent calls one interface instead of maintaining per-system code.
Every provider uses different auth flows — OAuth 2.0 for Greenhouse, API keys for Typeform, SOAP tokens for Workday. Managing credentials per tenant across dozens of systems is a full-time engineering burden. A managed auth layer handles token refresh, rotation, and secure storage automatically across all connected providers.
Without a search-first approach, the agent pre-loads every ATS and messaging action definition into its context window, burning tokens at scale. A tool-discovery layer lets the agent search by natural language and load only the actions it needs per request, keeping context lean and costs predictable.
Coverage gaps stall deployment at every new customer. StackOne's Connector Studio lets teams build and publish custom connectors for REST, SOAP, or GraphQL APIs in hours — covering long-tail ATS platforms and form tools without waiting on a vendor roadmap.
Referee responses are external, uncontrolled text fed directly into the agent's context. An attacker can embed instructions that trick the agent into changing a candidate's status or leaking data. A prompt injection guard that screens inbound data before the agent processes it mitigates this risk.
Referee submissions contain names, employment history, and candid performance assessments — all PII. The agent infrastructure should enforce scoped permissions so the agent only accesses fields it needs, and use a no-data-storage proxy model where responses are processed in transit without persisting raw PII in third-party systems.

Connect Your Agent to Your Recruiting Stack

Start building in minutes. MCP connectors to every system your agent needs.