Find completed interviews
Show me the candidates who completed the interview.
Get the list of candidates who have finished, without filtering the dashboard first. Useful as a start-of-day check on what is ready for review.
InterviewFlowAI MCP
Turn Your AI Assistant Into a Recruiting Assistant
Ask your recruiting data instead of digging through it.
Connect InterviewFlowAI to Claude, ChatGPT, Codex, or another MCP-compatible assistant and ask questions about your real candidates and interviews. No exporting, no copy-pasting candidate records into a chat window.
Ask things like:
- Show me the candidates who completed the interview.
- Show me the top 10 candidates by interview score.
- Show me the candidates with an interview score of 70 or above.
Already using InterviewFlowAI? Connect MCP in your workspace
Included on Growth, Scale, and Enterprise. Works with Claude, ChatGPT, Codex, and other MCP-compatible assistants.
You
Show me the top 10 candidates by interview score.
Called InterviewFlowAI · list candidates by score
CandidateRoleScore
Priya RaghunathanSupport Lead92
Daniel OkaforSupport Lead89
Mei Lin ChowSupport Lead86
Tomas AlvarezSupport Lead84
Hannah BergstromSupport Lead81
+ 5 more · shortlist ready for recruiter review
Works with your AI assistant
Claude DesktopRead and scoped write
Claude CodeRead and scoped write
ChatGPTRead and scoped write
Codex CLIRead and scoped write
Codex DesktopRead and scoped write
Ask, get real candidates back, decide as a team
Three views of the same workflow: a recruiter asks a question, InterviewFlowAI returns the candidates and scores behind it, and the hiring team decides who moves forward.
Your recruiting data should be as easy to use as asking a question.
You already use an AI assistant to draft job descriptions, rewrite outreach, and clean up interview notes. The moment the question involves your actual pipeline, the assistant goes quiet — it has no idea who applied, who finished an interview, or how anyone scored.
So you do the work yourself. Open the dashboard. Filter. Open a candidate. Open the interview. Read the report. Go back. Repeat for the next nine candidates. Then paste a few of them into a chat window to get help comparing. InterviewFlowAI MCP removes that step: your assistant asks InterviewFlowAI for the data it needs and answers from your workspace. It is the same shift we describe in AI candidate screening.
Generic AI assistant
- Recruiter
- Open dashboard
- Filter candidates
- Open each candidate
- Copy the details by hand
- Paste into AI assistant
- Answer
AI assistant + InterviewFlowAI MCP
- Recruiter
- Ask the AI assistant
- InterviewFlowAI
- Candidate + interview data
- Answer
Ask InterviewFlowAI in plain English
These are documented InterviewFlowAI MCP workflows, not concepts. Each card shows a prompt you can type today and what comes back.
Build a shortlist
Show me the top 10 candidates by interview score.
Returns a ranked list you can use as a starting point for recruiter review. You decide who moves forward.
Apply a score threshold
Show me the candidates with an interview score of 70 or above.
Triage a large pipeline down to the group worth your attention, then work through it in the assistant or open the full reports in InterviewFlowAI.
Review a candidate before a call
Summarize this candidate’s interview.
Pull a candidate’s interview into a summary you can read in 30 seconds before a hiring manager sync — instead of scanning a full transcript.
Update a candidate record
Add a note to this candidate: strong on stakeholder examples, needs a technical follow-up.
Workspace owners can update candidate records through the assistant. Confirm every change before you let the assistant make it.
What is MCP?
MCP (Model Context Protocol) is an open standard that lets AI assistants connect to outside tools and data sources. InterviewFlowAI MCP uses it to give a compatible assistant controlled access to your InterviewFlowAI workspace, so it can look up candidate and interview information and, where your role permits, make supported updates.
Practically speaking, MCP is the reason your assistant can answer a question about your pipeline instead of guessing. You add InterviewFlowAI once in your AI client, sign in with your InterviewFlowAI email, and the assistant can request data from your workspace when a question needs it.
You do not need to understand the protocol to use it. If you can add a connector in Claude Desktop, you can set this up. New to this? Read how to use an AI agent for recruiting.
1
Recruiter
2
Claude / ChatGPT / Codex / MCP-compatible assistant
3
InterviewFlowAI MCP
4
Your candidates + interviews
Use InterviewFlowAI from the AI tools you already work in
InterviewFlowAI MCP works with Claude, ChatGPT, and Codex.
| Client | Setup |
|---|---|
| Claude Desktop | Add a custom connector using the InterviewFlowAI server URL |
| Claude Code (CLI) | One claude mcp add-json command, then authenticate with the /mcp command |
| ChatGPT | Add InterviewFlowAI as a custom connector using the same server URL |
| Codex CLI | Add the InterviewFlowAI server to config.toml, then codex mcp login |
| Codex Desktop | Uses the same Codex configuration file, then authenticate in Settings → MCP Servers |
Any other MCP-compatible assistant can connect using the same server URL. Setup differs slightly per client, so follow the instructions for yours.
MCP server URL
https://api.interviewflowai.com/mcpConnect in minutes
Add InterviewFlowAI to your AI client
Open MCP in InterviewFlowAI and add the server URL to Claude, ChatGPT, Codex, or another MCP-compatible assistant. Available on Growth, Scale, and Enterprise plans.
Sign in
Authenticate with the same email address you use for your InterviewFlowAI account. That is what links the session to the right company workspace.
Ask about your pipeline
Start with a documented prompt: “Show me the candidates who completed the interview.” If you get your candidates back, you are connected.
Keep review and decisions with your team
Read-only questions are the safe place to start. Add write actions later, and confirm each one before the assistant makes a change.
Connect InterviewFlowAI
Prefer the exact commands? Read the MCP documentation
You stay in control
You stay in control of your candidate data.
Read and write are separate scopes, and the role you already have in InterviewFlowAI carries over to the assistant.
Recruiting data is sensitive. Confirm every write action before you allow your assistant to change a record.
Your own account, your own access
Authentication is tied to your InterviewFlowAI account through our identity provider. Sign in with the same email you use for InterviewFlowAI.
Read access is separate from write access
The read scope lets an assistant list and inspect AI Interviewers and Candidates. Nothing more.
Write access is scoped and role-restricted
The write scope lets a workspace owner update candidate records. InterviewFlowAI does not expose write tools to Viewers.
Confirm before anything changes
Review and approve every write action before you let your assistant apply it.
Hiring decisions stay with people
MCP helps you find, review, and organize candidate information. It does not decide who advances.
Built for real recruiting workflows
Written for recruiters, talent acquisition leaders, recruiting operations, and talent acquisition teams.
Candidate review
Show me the candidates who completed the interview.
Start your review from a list, not from a dashboard filter. See who is actually ready to be looked at.
Shortlist preparation
Show me the top 10 candidates by interview score.
Surface high-scoring candidates for recruiter review and hand a hiring manager a shortlist with the evidence behind it.
Automated ranking →Interview review before a debrief
Summarize this candidate’s interview.
Walk into a hiring manager conversation with the substance of the interview instead of a score.
Scorecards and transcripts →Recruiting operations
Add a note to this candidate and update their custom fields.
Workspace owners can keep candidate records current from the assistant, with each change confirmed first.
High-volume hiring
Show me the candidates with an interview score of 70 or above.
When a role has hundreds of completed interviews, one question narrows the pool to the group worth reviewing.
High-volume hiring →An AI assistant is far more useful when it can reach your recruiting data
The gap between a general assistant and a connected one is not the model. It is whether it can see your pipeline.
The difference is not a better model. It is a model that can see your pipeline — the same idea behind Ask FlowAI inside the product.
Where MCP fits
Bring AI interviews into your agentic recruiting workflow.
InterviewFlowAI already sits between your ATS and your first-round interviews. It integrates with Greenhouse, Ashby, Teamtailor, Zapier, and custom systems through our API, so interviews get triggered from the workflow your team already runs and results come back to the candidate record.
MCP adds a new way to work with what comes out of those interviews. It is an interaction layer over your InterviewFlowAI data — not a replacement for your system of record.
ATS / recruiting workflow
InterviewFlowAI
AI interviews
Candidate + interview data
InterviewFlowAI MCP
Claude / ChatGPT / Codex / MCP-compatible assistant
Frequently asked questions
What is InterviewFlowAI MCP?
InterviewFlowAI MCP is a connection that lets an MCP-compatible AI assistant work with your InterviewFlowAI candidate and interview data. Once connected, you can ask your assistant questions like “show me the candidates who completed the interview” and get the answer from your own workspace.
What can I do with InterviewFlowAI MCP?
You can list your AI Interviewers, see which candidates completed interviews, rank candidates by interview score, filter candidates by a score threshold, review candidate details, and summarize interviews. Workspace owners can also update candidate records.
Which AI assistants work with InterviewFlowAI MCP?
Claude, ChatGPT, and Codex all connect, including Claude Desktop, Claude Code, Codex CLI, and Codex Desktop. Any other MCP-compatible assistant can use the same InterviewFlowAI MCP server URL.
Can I use InterviewFlowAI MCP with Claude?
Yes. Claude Code and Claude Desktop are both documented. In Claude Desktop you add InterviewFlowAI as a custom connector using the InterviewFlowAI MCP server URL, then authenticate with your InterviewFlowAI email. In Claude Code you add the server with a single command and authenticate with the /mcp command.
Can I use InterviewFlowAI MCP with Codex?
Yes. Both Codex CLI and Codex Desktop are documented. You add InterviewFlowAI to your Codex configuration file and authenticate through your browser. Codex Desktop uses the same configuration file as the CLI.
Can I use InterviewFlowAI MCP with ChatGPT?
Yes. Add InterviewFlowAI as a custom connector in ChatGPT using the InterviewFlowAI MCP server URL, then sign in with the same email you use for InterviewFlowAI. Custom connector availability depends on your ChatGPT plan and workspace settings.
Can the AI assistant change candidate information?
Only with the write scope, and only for workspace owners. Workspace owners can update candidate records. InterviewFlowAI does not expose write tools to Viewers, and you should confirm every write action before allowing the assistant to make a change.
Does InterviewFlowAI MCP replace my ATS?
No. Your ATS stays your system of record. MCP is a way to interact with the candidate and interview data that InterviewFlowAI produces, using an AI assistant instead of the dashboard.
Can MCP make hiring decisions automatically?
No. InterviewFlowAI MCP helps you retrieve, review, and organize candidate information and surface candidates for recruiter review. Hiring decisions should stay under human review and follow your organization's own policies and legal obligations.
Do I need to be technical to set this up?
For Claude Desktop and ChatGPT, no. You fill in a connector form and sign in. Claude Code and Codex are command-line tools, so those setups involve pasting one command or one configuration block. Each path is documented step by step.
Which plans include InterviewFlowAI MCP?
MCP is included on the Growth, Scale, and Enterprise plans at no extra cost. It is not available on Pilot. See the InterviewFlowAI pricing page for plan details.
How this was verified
Capabilities described on this page were verified against the published InterviewFlowAI MCP documentation at docs.interviewflowai.com/platform/mcp in September 2026. Supported clients, permission scopes, and role restrictions can change; the documentation is the source of truth.
Ask your recruiting data instead of digging through it.
Connect InterviewFlowAI to Claude, ChatGPT, Codex, or another MCP-compatible assistant and get your first answer in a few minutes. Start with a read-only question and go from there.
New to AI recruiting agents? Read how to use an AI agent for recruiting →