Connect Lovelio to your AI agent

Ask your agent questions about your desk and get instant answers, or hand it the admin: create jobs, move candidates, send shortlists to clients, log placements. Ask in plain English and it works directly in your Lovelio account. Any MCP client works - Lovelio is not tied to one AI provider.

Your region:Your workspace lives in one region - it is the domain you sign in on. Every URL below updates to match.

What is MCP?

MCP stands for Model Context Protocol. It is an open standard, not one company’s product. Think of it as a secure bridge between an AI agent and your Lovelio account. Once connected, the agent can look up your clients, jobs, candidates, and applications whenever you ask - and, if your key allows it, take action on them too.

It works like giving a trusted assistant a login. Approve read permissions and the agent can only look things up. Approve write permissions and it can act: create a job, move a candidate, send a shortlist to a client. Every action runs through the same permissions and activity logging as the app, so an agent can never do something you could not do yourself.

Because MCP is a standard, you choose the client. Any desktop app, code editor, or agent that speaks MCP connects to the same Lovelio server URL, and switching clients later changes nothing on our side.

You decide what the agent can do

The agent receives only the permissions you approve on the Lovelio consent page. No API key is shared with the client. You can disconnect at any time.

What you can do

“Find me senior engineers who applied this week”

Search across all candidates and applications using natural language

“How is the Product Manager role going?”

Get an AI summary of any job's pipeline health and top candidates

“Who are our strongest candidates right now?”

Find candidates by skills, experience, or assessment scores

“Show me the AI assessment for Sarah's application”

Pull up detailed scoring breakdowns and recommendations

“What interviews do we have scheduled?”

See upcoming interviews with format, status, and feedback

“Create a job for Acme: senior backend engineer in Leeds, 80k”

Drafts the spec, ad, and interview plan for you to review before it goes live

“Submit the top three candidates to the client”

Builds the submission, drafts the summaries, and sends the client a review link

“Log the placement for Sarah - 95k salary, 20% fee”

Records the win with fee, start date, and guarantee tracking

“What happened in hiring today?”

Review recent activity - applications, stage changes, emails, notes

“Give me an overview of our account”

See your plan, team, and account details

Getting started

Connecting takes about 2 minutes. Here is what you need:

Add the Lovelio server to your AI client

Every MCP client has a place to add a server. Paste this URL into it: https://us.lovelio.ai/api/mcp. Desktop apps usually call it Connectors or Custom connector; editors and command-line agents take a config file or an add command. Refer to Technical setup below for the config block.

Approve the requested permissions

A Lovelio consent page opens. Check the client and its requested permissions, clear any you do not need, then approve. Dynamically registered clients have not been reviewed by Lovelio.

Start asking questions

That is it. Ask your agent anything about your hiring data. Try: “How many open jobs do we have?”

Example conversations

You

Give me a morning update on our hiring. What happened yesterday and what needs attention today?

Your agent

It checks your recent activity, pipeline stats, and scheduled interviews to give you a summary like: “You received 12 new applications yesterday across 3 roles. The Senior Engineer role has 4 candidates in shortlist who haven’t been contacted yet. You have 2 interviews scheduled today...”

Available tools

Your agent gets 70 tools: 24 that read your Lovelio data and 46 that act on it. You do not need to know the tool names - ask in plain English and the agent picks the right tool automatically. Tools marked write only work when you grant the required write permission.

Search

searchNatural language search across jobs, candidates, applications, and interviews

search_candidatesSemantic search for candidates by skills, experience, or qualifications with match scores

Clients

list_clientsList the companies your agency recruits for, with industry and Client DNA status

get_clientFull client detail: website, industry, locations, enrichment, and notes

create_clientwriteCreate a client - pass a website and Lovelio enriches the rest automatically

update_clientwriteUpdate client fields like name, website, industry, and locations

add_client_contactwriteAdd a contact (hiring manager, HR, finance) to a client

update_client_contactwriteUpdate a contact's details on a client

add_client_dna_notewriteAdd a note or meeting notes to a client's hiring DNA

Jobs

list_jobsList jobs filtered by status, with location and application counts

get_jobFull job details including description, criteria, interview questions, and pipeline stats

get_job_summaryAI-generated recruitment summary with pipeline health and recommended actions

get_job_share_bundleReady-to-share content bundle: social posts, careers URL, and email summary

get_job_rankingStack-rank interviewed candidates on a job by scorecard score

create_jobwriteCreate a job requisition for a client - the description, criteria, and interview plan are AI-generated

create_job_from_descriptionwriteTurn a plain-language role description into a job draft for review

get_job_draftwriteRead the current state of an in-progress job draft, including remaining gaps

update_job_draftwriteFill gaps on a job draft, including the required client

revise_job_draftwriteRewrite a job draft from a natural-language instruction

confirm_job_draftwritePromote a reviewed draft into a real job

get_job_creation_statuswritePoll per-step progress while a new job is being generated

update_jobwriteUpdate job fields like title, location, compensation, and status

close_jobwriteClose a job as filled, closed, or on hold - live ads come offline

reopen_jobwriteReopen a closed or filled job

Job ads & social posts

list_job_adsList job advertisements with channels and publish status

get_job_adwriteFetch the most recent published ad for a job

update_job_adwriteEdit a job ad's title, description, or published flag

rewrite_job_adwriteAI-rewrite a job ad from a natural-language instruction

get_social_draftwriteRead a job's LinkedIn and X post drafts

update_social_draftwriteEdit a job's LinkedIn and X post text

Candidates

list_candidatesList candidates with skills and contact info

get_candidateFull candidate profile including resume data and CV link

create_candidatewriteCreate a candidate - pass a resume URL and the CV is parsed automatically

update_candidatewriteUpdate a candidate's profile, skills, and tags

Applications

list_applicationsList applications filtered by job, candidate, or pipeline stage

get_applicationFull application with AI assessment scores, criteria breakdown, and recommendation

move_stagewriteMove an application to a different pipeline stage

reject_candidatewriteReject an application

withdraw_applicationwriteMark an application as withdrawn by the candidate

add_notewriteAdd a note to any record - it shows in the activity timeline

Interviews

list_interviewsList interviews with schedule, format, and outcome

get_interviewFull interview details including feedback and notes

schedule_interviewwriteSchedule an internal screen or a client interview

reschedule_interviewwriteMove an interview to a new time

cancel_interviewwriteCancel a scheduled interview

update_interview_outcomewriteRecord the outcome and feedback for an interview

submit_scorecardwriteSubmit per-criterion scores and an overall recommendation

Submissions

list_submissionsList submissions sent to clients, with response counts

get_submissionOne submission in full: candidates, review link, and client verdicts

create_submissionwriteSend a submission of candidates to a client contact for review

chase_submissionwriteNudge a client contact who has not responded

answer_submission_questionwriteAnswer a client's question on a submitted candidate, or send them the review link again

Placements & offers

list_placementsList placements with candidate, client, fee, and start date

get_placementOne placement in full: salary, fee, start date, and guarantee

log_offerwriteRecord that the client made an offer (salary, offer date, start date) and move the application to the Offer stage

record_offer_outcomewriteRecord the offer's answer: accepted, declined (by candidate or client) or countered

create_placementwriteRecord a placement: candidate, client, job, salary, fee, and start date

mark_placement_statuswriteUpdate a placement through its guarantee lifecycle

Talent pools

create_talent_poolwriteCreate a named group for candidates who fit your agency

add_to_talent_poolwriteAdd a candidate to a talent pool

remove_from_talent_poolwriteRemove a candidate from a talent pool

delete_talent_poolwriteDelete a pool - the candidates themselves always stay

Comms

send_emailwriteEmail a candidate - outbound email must be switched on for the account first

Activity

list_activitiesAudit log of who did what and when - applications, emails, stage changes, notes

Account & team

get_account_overviewYour company name, plan, location, and account details

get_transcription_usageThis month's video-transcription usage against your plan

get_company_intelligenceYour synthesised company profile: voice, values, hiring process

refresh_company_intelligencewriteQueue a fresh company profile synthesis run

invite_team_memberwriteInvite a new team member by email

create_webhookwriteCreate a webhook subscription for your account

Batch

batchwriteRun up to 100 actions in one round-trip, each retried independently

Security & privacy

Your permissions, nothing more

The agent has exactly the permissions approved on the consent page. Every action follows the same access checks and activity logging as the app.

Encrypted storage

Your API key is encrypted with AES-256-GCM at rest. Access tokens are stored as SHA-256 hashes - the raw token is never saved.

Token expiry

Access tokens expire after 1 hour and are automatically refreshed. Refresh tokens expire after 30 days.

Disconnect anytime

Revoke the MCP credential in Settings > API keys to cut off access immediately. You can also disconnect in your MCP client.

Technical setup

For developers or advanced users who want to configure the connection manually.

Everything an MCP client needs is this one URL:

https://us.lovelio.ai/api/mcp

Clients that take a JSON config file want this block. Most use the key mcpServers; a few name it differently, so check your client’s docs for the wrapper.

{
  "mcpServers": {
    "lovelio": {
      "type": "http",
      "url": "https://us.lovelio.ai/api/mcp"
    }
  }
}

Clients with a settings screen ask for the same URL under a heading like Connectors, Custom connector, or Add MCP server. Clients with a command line have their own add command - pass the URL and the HTTP transport.

With OAuth, the first call opens a Lovelio consent page. Client callback support can vary. If your client supports HTTP headers, you can instead use a dedicated scoped API key from this region in its secure Authorization header settings. Company keys have agency-wide record access; personal OAuth uses the connected person’s role and record access.