Play Video

Oppora's MCP is Live for Claude and ChatGPT. Connect our Email Database & Outreach features to build smart automations inside your AI chat. Start for Free

Oppora MCP server

The Oppora MCP server is a remote Model Context Protocol endpoint for Claude, ChatGPT, Cursor, VS Code and Claude Code. Its 81 tools find companies, people and job posts, find and verify emails and phone numbers, and run outreach: email campaigns, multichannel sequences with LinkedIn steps, workflows and analytics, all inside your Oppora account.

81 toolsSign in with Oppora on Claude and ChatGPTCredits apply as in the app

Last updated:

Connector URL (Claude, ChatGPT)
https://api.oppora.ai/mcp/app

Sign in with your Oppora account and pick an organization. No API key.

API key endpoint (Cursor, VS Code, scripts)
https://api.oppora.ai/mcp

Send an Oppora API key as Authorization: Bearer or X-API-Key.

Both use Streamable HTTP. To disconnect Claude or ChatGPT, open Integrations > Oppora API in Oppora and click Disconnect on the connected app, or remove the connector in your assistant.

Quick start

1

Add the connector

In Claude or ChatGPT, add a custom connector with the URL https://api.oppora.ai/mcp/app. Using Cursor or VS Code? Create an API key instead.

Create an API key
2

Sign in

Sign in with your Oppora account, pick the organization to connect and click Allow. Steps for each client are below.

3

Try a tool

Ask "what lists do I have?" and the assistant calls my_lists. Then try one of the example prompts.

Connect a client

Claude.ai and Claude Desktop. Open Settings > Customize > Connectors, click +, choose Add custom connector, paste the URL below and click Add. Then click Connect on the Oppora connector. Claude's Free plan allows one custom connector.

https://api.oppora.ai/mcp/app

Sign in with your Oppora account, pick the organization to connect, then click Allow. No API key needed.

Team and Enterprise: an owner adds it under Organization settings > Connectors > Add > Custom > Web. Each member then clicks Connect under Customize > Connectors and signs in with their own Oppora account.

Connectors you add on claude.ai also appear in Claude Desktop when you sign in with the same account.

Example prompts

Paste any of these into a connected client. Under each one is the tool chain it usually triggers, so you can see where credits go before you run it.

  1. Find Series B software companies in Germany with 51 to 500 employees. Save them to a new list called "DACH Series B", then find their VP Sales and Heads of Sales.

    Tools: get_available_countries, get_available_last_funding_round_types, get_available_industries, discover_companies, create_list, add_companies_to_list, get_available_departments, get_available_management_levels, find_leads_for_companies, get_task_status

  2. Find the work email for Priya Shah at acme.com.

    Tools: email_search, get_job_status (only if the lookup is still running)

  3. Check these 20 addresses before I import them and tell me which will bounce. Run the advanced check on any that come back risky.

    Tools: email_verifier (standard, once per address), email_verifier (advanced, for risky ones)

  4. Find emails for everyone in my "DACH Series B" list, then tell me how many are verified.

    Tools: my_lists, find_emails_for_leads, get_task_status, get_list_analytics

  5. Build a draft sequence for "DACH Series B": a LinkedIn connection request, wait 2 days, an intro email, then a follow-up email 4 days later. Send from my Gmail mailbox.

    Tools: list_sequence_templates, create_sequence_campaign, save_sequence, list_email_accounts, attach_sequence_audience, get_sequence

  6. How is the "DACH Series B" sequence doing? Who replied, and who is stuck on a step?

    Tools: get_sequence_campaign_analytics, get_sequence_recipients

  7. Which companies in Texas posted SDR jobs this week? Save them to a list and find their sales leaders.

    Tools: search_jobs, create_list, add_jobs_to_list, find_leads_for_companies, get_task_status

  8. Find 25 companies that match my ICP and add them to my "Q4 targets" list.

    Tools: get_icp_profile, get_available_industries, get_available_countries, discover_companies, my_lists, add_companies_to_list

What you can do that a data-only MCP can't

A data MCP server answers lookups. Oppora's server also acts on the results in the same account, with the same credits and the same lists you see in the app.

Find and verify

  • Search companies, people and live LinkedIn job posts.discover_companies, discover_people, search_jobs
  • Find work emails through Oppora's email waterfall, then any finder you connected with your own key.email_search, find_emails_for_leads
  • Find phone numbers through a phone waterfall. Numbers already in Oppora cost nothing.phone_search
  • Verify deliverability, with a catch-all aware advanced mode.email_verifier
  • Fill firmographic fields and score leads against your offer.enrich_companies, score_leads

Act on it

  • Set up email campaigns with templates, follow-ups, schedules and mailboxes, check readiness and start them.create_or_update_campaign, check_campaign_readiness, start_campaign
  • Build multichannel sequences: email, LinkedIn invites, messages, InMail, profile visits, likes, comments and follows, plus waits, conditions, call tasks and CRM pushes.save_sequence, attach_sequence_audience, activate_sequence
  • Report on campaigns, sequence steps, each recipient, mailbox health and list quality.get_sequence_campaign_analytics, get_sequence_recipients, get_email_account_health
  • Read and update your ICP, and run repeatable workflows that target it.get_icp_profile, update_icp_profile, create_workflow, set_workflow_icp

Tools reference

81 tools

Every tool returns TOON-encoded text except discover_companies and discover_people, which return JSON. Open a row for arguments, response shape and an example.

81

Discovery (companies & people)

5

Search Oppora's company and people databases. Rows carry an Oppora id you can pass to the add-to-list tools. A call that returns results costs a flat 5 data credits for companies or 10 for people; empty results are free. Reuse next_cursor for the next page.

discover_companiesSearch companies by industry, size, location, revenue, funding, growth and tech stack.
5 data credits per call
Arguments
organization_name?: string | string[], industry?: string[], category?: string[], location?: string[] (city or state), size?: string[] ("2-10" … "10001+"), hq_country?: string[] (ISO 3-alpha), company_type?: string[], year_founded?: string, revenue_range_min/max?: int (USD), total_funding_range_min/max?: int, latest_funding_amount_range_min/max?: int, last_funding_round_type?: string[], last_funding_date?: string, employee_growth_6m_min / employee_growth_12m_min?: int (percent), acquisition_status?: "acquired" | "not_acquired", technologies?: string[], department_headcount?: {department, min?, max?}[], department_headcount_growth?: {department, min?, max?, period?: "6m" | "12m"}[], company_website_domain?: string[], domain?: string, domain_pattern?: string, linkedin_url?: string[], exclude_companies / exclude_industries / exclude_keywords / exclude_domains / exclude_size?: string[], limit?: int (1-100, default 25), next_cursor?: string
Returns

JSON { count, credit_charged, credits_remaining, next_cursor, has_more, results: [{ id, name, domain, website, linkedin_url, industry, employee_count, location, country, … , oppora_url }] }. id is set when the company is already in your Oppora database.

Example
discover_companies(
  hq_country=["USA"],
  industry=["Software Development"],
  size=["51-200", "201-500"],
  revenue_range_min=5_000_000,
  technologies=["Salesforce"],
  limit=25,
)
discover_peopleSearch people by title, department, seniority, company, location and skills.
10 data credits per call
Arguments
title?: string | string[], first_name?: string, last_name?: string, company_name / company_linkedin / company_domain?: string | string[], company_industries?: string[], company_hq_location?: string[], departments?: string[], management_levels?: string[], location?: string | string[], employee_count?: string | string[], years_experience?: string | string[], skills?: string[], recently_changed_jobs?: bool, open_to_cards?: ("CAREER_INTEREST" | "HIRING_MANAGER" | "VOLUNTEERING")[] (open to work / hiring / volunteering), exclude_companies / exclude_industries / exclude_titles / exclude_keywords / exclude_location?: string[], limit?: int (1-100, default 25), next_cursor?: string
Returns

JSON { count, has_more, next_cursor, credit_charged, credits_remaining, results: [{ first_name, last_name, title, department, management_level, company_name, company_domain, location, linkedin_url, … }] }. Running the same search again without next_cursor returns the same rows and is charged again.

Example
discover_people(
  title=["VP Sales", "Head of Sales"],
  management_levels=["VP"],
  company_domain=["stripe.com", "plaid.com"],
  limit=25,
)
enrich_companyLook up one company by LinkedIn URL or domain: firmographics, headcount growth, tech stack, open jobs, web traffic, decision makers and news. Saves it to Oppora.
10 data credits per company found, 20 with tech stack
Arguments
linkedin_url?: string, domain?: string (one is required; linkedin_url wins when both are given)
Returns

TOON { company{ id, name, domain, website, linkedin_url, industry, size, employee_count, location, country, revenue_estimate_low_usd, revenue_estimate_high_usd, total_funding_usd, last_funding_round_type, last_funding_date, employee_growth_6m, employee_growth_12m, description, …, gtm{ tech_stack{ total, top[] }, hiring{ open_jobs, openings_growth_percent, recent_titles }, web_traffic{ monthly_visitors, … }, decision_makers[], cxos[], founders[], news[] } }, credit_charged, credits_remaining }. Needs 20 data credits available. No match is free.

Example
enrich_company(domain="stripe.com")
enrich_personLook up one person by LinkedIn profile URL: title, department, seniority, time in role, summary, skills, education, followers, employment and current company. Saves them to Oppora.
5 data credits per profile found
Arguments
linkedin_url: string
Returns

TOON { person{ linkedin_url, name, first_name, last_name, headline, current_title, location, summary, skills[], years_of_experience, education[], lead_id, company{ id, name, domain, industry, employee_size, location }, employment{ current[], past[] } }, credit_charged, credits_remaining }. No email; use email_search with the returned name and company domain. No match is free.

Example
enrich_person(linkedin_url="https://www.linkedin.com/in/jane-doe")
live_people_searchSearch people on LinkedIn in real time: current titles, companies and locations as they are today, not from a stored database.
6 data credits per page of 25
Arguments
title?, location?, current_companies?, past_companies?, past_titles?, schools?, skills?: string[]; company_name?, first_name?, last_name?: string; departments?, management_levels?, company_industries?, employee_count?, company_hq_location?, years_experience?, years_at_company?: string[]; recently_changed_jobs?, posted_on_linkedin?: bool; exclude_titles?, exclude_current_companies?, exclude_locations?: string[]; page?: int
Returns

TOON { count, total, page, has_more, next_page, credit_charged, credits_remaining, results: [{ name, headline, title, company, company_linkedin, location, linkedin_url }] }. Empty pages are free. Free plans get a short daily preview.

Example
live_people_search(title=["VP Sales"], location=["Austin, Texas"], recently_changed_jobs=True)

LinkedIn activity (live)

5

Live LinkedIn posts, comments and reactions for personalisation and intent: what a prospect posts or reacts to, what a company announces, who is posting about a topic, and who engaged with a post. Each page holds up to 100 items and costs 1 data credit when it returns results; empty pages are free. Pass next_page and next_page_token back to get more.

linkedin_profile_activityA person's posts, the comments they wrote, or the posts they reacted to. kind: posts, comments or reactions.
1 data credit per page
linkedin_company_postsA company's posts: launches, hiring, funding and events.
1 data credit per page
linkedin_search_postsFind posts by keywords, author, company, employees of a company, or company mentions.
1 data credit per page
linkedin_get_postRead one post by URL: full text, author, date and engagement.
3 data credits per post
linkedin_post_engagementWho commented on or reacted to a post, or the replies and reactions on a comment. Turns a post into a warm lead list.
1 data credit per page

Jobs

2

The Job Finder over MCP: live LinkedIn job posts, filtered by title, place, date and hiring company size. Saving posts to a list also adds each hiring company.

search_jobsSearch live LinkedIn job posts. Empty pages are free and nothing is saved.
1 data credit per page
Arguments
job_title?: string[], location?: string[], posted_limit?: "24h" | "week" | "month", company_size?: string[] (buckets to keep), workplace_type?: ("office" | "hybrid" | "remote")[], employment_type?: string[], experience_level?: string[], industry?: string[], job_function?: string[], salary?: "40k+" … "200k+", under_10_applicants?: bool, easy_apply?: bool, sort_by?: "date" | "relevance", next_cursor?: string
Returns

TOON { count, total, credit_charged, credits_remaining, next_cursor, has_more, jobs[N]{ job_title, company, company_id, company_size, job_location, posted_date, job_url, company_linkedin } }

Example
search_jobs(
  job_title=["SDR", "Sales Development Representative"],
  location=["Texas"],
  posted_limit="week",
)
add_jobs_to_listSave job rows from search_jobs to a list and add each hiring company to it.
Free
Arguments
list_id: int, jobs: list[dict] (1-100 rows from search_jobs), skip_duplicate?: bool (default true: one job per company)
Returns

TOON { added_jobs, added_companies, skipped, message, list_id, list_url }

Filter helpers

7

Return the exact values the discovery and lead tools accept. Misspelled values match nothing, so assistants call these first. All free.

get_available_industriesAbout 435 LinkedIn industry names.
Free
get_available_countriesCountry names with ISO 3-alpha codes. Pass iso_alpha3 to hq_country.
Free
get_available_company_typesPrivately Held, Public Company, Nonprofit and the rest.
Free
get_available_last_funding_round_typesFunding rounds such as seed, series_a, series_b.
Free
get_available_departmentsValues for departments on discover_people and find_leads_for_companies.
Free
get_available_management_levelsSeniority values such as C-Level, VP, Director, Manager.
Free
search_locationsAutocomplete up to 20 city and region names for location filters.
Free

Saved companies & leads

8

Manage records saved in your Oppora lists.

my_listsYour lists with company and lead counts. The only source of a valid list_id.
Free
create_listCreate a list, or return the existing one with that name (case-insensitive).
Free
Arguments
name: string (max 200 characters)
Returns

TOON { id, name, is_existing, message, list_url }

Example
create_list(name="Top Prospects Q4")
list_my_saved_companiesSaved companies, optionally for one list, filtered by industry and size.
Free
list_my_saved_leadsSaved leads, filterable by list and lead status.
Free
add_companies_to_listAdd Oppora company IDs to a list. Companies already in it are skipped free.
1 data credit per new company
remove_companies_from_listRemove companies from a list. The company records stay.
Free
add_leads_to_listAdd Oppora lead IDs to a list. Leads already in it are skipped free.
1 data credit per new lead
remove_leads_from_listRemove leads from a list. The lead records stay.
Free

Bulk upsert (external data)

2

For data that did not come from discovery: web scrapes, news results, pasted lists, CSV rows. Both tools dedupe against Oppora's database and can add the rows to a list in the same call. Saving is free.

create_companies_bulkUpsert up to 100 companies. Dedupes by name, LinkedIn, website and domain.
Free
Arguments
companies: list[{ name (required), domain?, website?, linkedin?, industry?, approx_employee_size?, meta? }], list_id?: int
Returns

TOON { company_ids, added_to_list, list_id, list_url, message }

Example
create_companies_bulk(
  companies=[
    {"name": "Acme", "domain": "acme.com"},
    {"name": "Globex", "linkedin": "https://linkedin.com/company/globex"},
  ],
  list_id=12,  # optional: save and add to the list in one call
)
create_leads_bulkUpsert up to 100 leads. Dedupes by email, then LinkedIn, then name plus company.
Free
Arguments
leads: list[{ first_name or email (one required), last_name?, linkedin?, title?, phone?, location?, company_name?, company_domain?, company_linkedin?, company_industry? }], list_id?: int. With list_id, rows also accept email_status, ai_recommendation, status, alternate_emails, source, custom_columns.
Returns

TOON { created, existing, failed, lead_ids, failed_details, list_id, message }

Example
create_leads_bulk(
  leads=[
    {"first_name": "Alice", "email": "[email protected]", "company_name": "Stripe"},
    {"linkedin": "https://linkedin.com/in/bob-smith"},
  ],
  list_id=12,
)

Async list operations

4

Long-running work on a list. Each call returns a task_id at once and runs in the background; poll it with get_task_status. Every response includes an oppora_url for the list.

enrich_companiesFill firmographic fields (website, industry, size, description and more) for companies in a list.
1 data credit per field filled
Arguments
list_id: int, company_ids?: int[] (omit for the whole list)
Returns

TOON { task_id, message, list_id, list_name, counts, oppora_url }

Example
enrich_companies(list_id=12)
find_leads_for_companiesFind people at the companies in a list from Standard, Premium or Deep sources.
1/3/5 data credits per lead
Arguments
list_id: int, departments: string[], management_level: string[], company_ids?: int[], title?: string[], lead_limit?: int (per company, default 10, max 1000), country / country_full_name / location / experience / employee_count?: string[], source_tiers?: ("standard" | "premium" | "deep")[], exclude_location?: string[], main_objective?: string
Returns

TOON { task_id, message, list_id, list_name, oppora_url }

Example
find_leads_for_companies(
  list_id=12,
  departments=["Sales"],          # from get_available_departments
  management_level=["VP", "Director"],
  title=["VP Sales", "Head of Sales", "Sales Director"],
  lead_limit=5,
  source_tiers=["standard"],
)
score_leadsScore each lead in a list Yes or No against your offer, with a reason. Fills gaps from your ICP.
1 AI credit per lead
Arguments
list_id: int, service_type: string, leads_per_company: int, priority_departments: string, specific_leads_per_department: string, additional_requirements?: string
Returns

TOON { task_id, message, list_id, list_name, counts, oppora_url }

Example
score_leads(
  list_id=12,
  service_type="Sales automation software for outbound teams",
  leads_per_company=3,
  priority_departments="Sales, Marketing",
  specific_leads_per_department="VP Sales x2, Director of Marketing x1",
)
find_emails_for_leadsFind work emails for leads in a list through the email waterfall. Misses are free.
Data credits per email found
Arguments
list_id: int, lead_ids?: int[] (omit for the whole list)
Returns

TOON { task_id, message, list_id, list_name, counts, oppora_url }

Example
find_emails_for_leads(list_id=12)

Tasks (poll async work)

2

Read background tasks started by the list operations and workflows. Each row links to the affected list or workflow.

get_task_statusStatus, progress, summary, credits used and oppora_url for one task.
Free
Arguments
task_id: int
Returns

TOON { task_id, title, status, progress, summary, credits_used, oppora_url, … }

Example
get_task_status(task_id=1408)
list_my_tasksYour recent tasks, newest first, filterable by status.
Free
Arguments
status?: "pending" | "in_progress" | "completed" | "stopped", page?: int, limit?: int (max 100)
Returns

TOON { total, page, limit, has_more, tasks[N]{ task_id, title, status, progress, summary, oppora_url, … } }

Example
list_my_tasks(status="in_progress", limit=10)

Workflows

3

Chain steps (find companies, enrich, find leads, score, find and verify emails, draft a campaign) into a workflow on a list. Run it now or at a scheduled time.

create_workflowCreate a workflow on a list, with an optional one-time schedule.
Free
Arguments
name: string, list_id: int, scheduled_time?: "YYYY-MM-DD HH:MM", timezone_str?: string (IANA, default "US/Eastern")
Returns

TOON { workflow_id, message, oppora_url }

Example
create_workflow(name="Weekly outbound", list_id=12)
add_workflow_nodeAppend one or more steps to a workflow, in run order.
Free
Arguments
workflow_id: int, list_id: int, prompt: string (step types plus instructions)
Returns

TOON { workflow_id, message, step_count, pipeline, oppora_url }. Step types: company-crawl, company-enrich / ai-enrich, company-filter, job-finder, lead-search, lead-enrichment, ai-recommendation, lead-remover, search-contact, mobile-number, verify-contact, generate-campaign, start-campaign, crm-write, web-search, delay.

Example
add_workflow_node(
  workflow_id=79,
  list_id=12,
  prompt="company-crawl: NYC HR firms, lead-search: HR directors, search-contact, verify-contact",
)
start_workflowRun a workflow now. Steps run in order in the background; track them with list_my_tasks.
Free (steps bill their own credits)
Arguments
workflow_id: int, list_id: int
Returns

TOON { message, oppora_url }

Example
start_workflow(workflow_id=79, list_id=12)

ICP

3

Read and update your organization's ideal customer profile, and point workflow steps at it. Generating a new ICP happens in the Oppora app.

get_icp_profileYour ICP: target companies, buyer persona, messaging and signals, with a mapping to discovery filters.
Free
Returns

TOON { status, company_context, target_company, buyer_persona, messaging, signals, summary, confidence_score, refining, can_edit, message, oppora_url }

update_icp_profileChange ICP sections. Owner or admin only; lists are replaced, not appended.
Free
Arguments
company_context?, target_company?, buyer_persona?, messaging?, signals?: dict, summary?: string (pass at least one)
Returns

TOON { status, updated_fields, message, oppora_url }

set_workflow_icpTurn "use my ICP" on or off for a workflow's steps.
Free
Arguments
workflow_id: int, list_id: int, use_icp?: bool (default true), steps?: string[] (omit for every step that supports it)
Returns

TOON { message, pipeline, icp_warning, workflow_id, list_id, oppora_url }

Example
set_workflow_icp(workflow_id=79, list_id=12)

Contact resolution

4

One-off finders and a verifier. Credits are charged only when a result comes back. A lookup that outlasts the call returns status: "pending" with a job_id; the assistant reads it with get_job_status instead of searching again.

email_searchFind a work email from a name and company domain via Oppora's waterfall, then your own connected finders.
1 data credit when found
Arguments
first_name?: string, last_name?: string, full_name?: string, domain: string (required)
Returns

TOON { email, status, source, credit_charged, credits_remaining }. Charged only when status is valid or risky.

Example
email_search(first_name="Alice", last_name="Chen", domain="stripe.com")
email_verifierCheck whether an address will bounce. Advanced mode adds a catch-all aware deep check.
1 data credit, 3 advanced
Arguments
email: string, mode?: "standard" | "advanced"
Returns

TOON { email, mode, status, credit_charged, credits_remaining } with status valid, invalid, risky, catch-all or unknown.

Example
email_verifier(email="[email protected]", mode="standard")
phone_searchFind a phone number from a LinkedIn URL (best) or name plus company or domain. Numbers already in Oppora are free.
1 phone credit when found
Arguments
linkedin_url?: string, first_name?: string, last_name?: string, full_name?: string, company?: string, domain?: string
Returns

TOON { phone, status, source, credit_charged, credits_remaining }

Example
phone_search(linkedin_url="https://linkedin.com/in/alice-chen")
get_job_statusRead a pending email_search, email_verifier or phone_search lookup.
Free
Arguments
job_id: int
Returns

TOON { id, kind, status, total, processed, credits_charged, error, results[] }

Campaigns

8

Email campaigns: inspect, set up, check readiness and start. Starting runs the readiness check and refuses if setup is incomplete.

list_campaignsCampaigns with sent, opened, clicked, replied and bounced counts and rates.
Free
get_campaignFull setup of one campaign: lists, mailboxes, templates, schedule, stats.
Free
get_campaign_reportFunnel for one campaign over an optional date range.
Free
create_or_update_campaignCreate or update a campaign: lists, templates, follow-ups, schedule, mailboxes, limits, tracking.
Free
Arguments
campaign_id?: int (omit to create), name?: string, status?: string, lists?: int[], template_id?: int, linkedin_template_id?: int, follow_up_templates_ids?: int[], schedule_id?: int, accounts?: int[], start_time / end_time?: "YYYY-MM-DD", daily_limit?: int, min/max_time_gap_between_emails?: int (minutes), sending_priority?: string, send_linkedin_invite, open_tracking, click_tracking, stop_tracking_on_reply, is_sent_invalid_emails, delivery_optimization, unsubscribe_header_enabled, skip_dnc_list, skip_rejected_domains, skip_rejected_domains_per_account, send_to_all_emails?: bool
Returns

TOON { campaign_id, name, status, created, updated, oppora_url, setup_url, message }. List-type fields replace the current values.

update_campaign_statusSet draft, active, paused or completed. Active runs the readiness check first.
Free
start_campaignStart a campaign. Refuses and lists the issues if setup is incomplete.
Free
check_campaign_readinessEvery missing setup piece, each with a link to fix it.
Free
list_email_accountsConnected Gmail, Outlook and SMTP mailboxes with status and today's sends.
Free

Sequences (multichannel)

9

Sequences mix email and LinkedIn steps with waits, conditions, call tasks and CRM pushes. New sequences start as drafts, and the editing tools refuse while a sequence is sending. activate_sequence and resume_sequence send real email and LinkedIn messages; their instructions limit them to an explicit request from you.

list_sequence_templatesReady-made sequence templates to seed a new campaign from.
Free
Arguments
include_steps?: bool (default false)
Returns

TOON { count, templates[N]{ key, name, flow, … } }

create_sequence_campaignCreate a sequence as a draft, empty or from a template. Contacts nobody.
Free
Arguments
name: string, template_key?: string
Returns

TOON { campaign_id, name, status, nodes, revision, requirements, next_step, oppora_url }

Example
create_sequence_campaign(name="DACH Series B")
get_sequenceSteps, settings, audience, and the errors and warnings that affect launch.
Free
Arguments
campaign_id: int
Returns

TOON { campaign_id, name, status, nodes, revision, requirements, errors, warnings, enrolled_leads, oppora_url }

save_sequenceReplace a draft's steps. AI comment and auto-reaction steps cost 1 AI credit per enrolled person.
Free to save
Arguments
campaign_id: int, nodes: list (the complete step graph; steps left out are deleted). action_type: send_email, send_connection, send_message, send_inmail, send_follow_up, visit_profile, like_post, comment_post, follow_user, withdraw_request, wait, condition, manual_call, push_to_crm, ai_reply
Returns

TOON { campaign_id, nodes, revision, warnings, requirements, paused, oppora_url }

attach_sequence_audienceSet the lead lists, mailboxes and LinkedIn accounts for a draft. Each argument replaces the current set.
Free
Arguments
campaign_id: int, list_ids?: int[], email_account_ids?: int[], linkedin_account_ids?: int[]
Returns

TOON { campaign_id, lists, email_accounts, linkedin_accounts, connected_linkedin_accounts, lead_count, requirements, oppora_url }

update_sequence_settingsChange a draft's name, schedule, dates, daily limit, tracking, CRM push and reply agent.
Free
Arguments
campaign_id: int, name?: string, schedule_id?: int, start_date / end_date?: "YYYY-MM-DD", daily_limit?: int (1-2000), push_replies_to_crm, stop_tracking_on_reply, open_tracking, click_tracking, skip_dnc, reply_agent_enabled?: bool
Returns

TOON { campaign_id, updated[], oppora_url }

activate_sequenceLaunch a sequence. Sends real email and LinkedIn messages; refused while readiness blockers remain.
Free
Arguments
campaign_id: int
Returns

TOON { campaign_id, name, status, enrolled, step_count, warnings, readiness_warnings, oppora_url }

pause_sequencePause a running sequence. Further sends stop at once.
Free
Arguments
campaign_id: int
Returns

TOON { campaign_id, status, oppora_url }

resume_sequenceResume a paused sequence after a fresh readiness check.
Free
Arguments
campaign_id: int
Returns

TOON { campaign_id, status, readiness_warnings, oppora_url }

Templates & schedules

7

Reusable pieces for email campaigns.

list_templatesEmail and LinkedIn templates, filterable by kind.
Free
list_follow_up_templatesFollow-up templates ordered by position and delay.
Free
list_schedulesSending windows with timezone and days.
Free
create_email_templateHTML email template with merge variables and fallbacks such as {industry|your industry}.
Free
create_linkedin_templatePlain-text LinkedIn message, 200 characters max.
Free
create_follow_up_templateFollow-up with position, days after the first email and an optional subject.
Free
create_scheduleSending window: days, start and end time, any IANA timezone.
Free

Analytics

7

Read-only reporting across campaigns, sequences, mailboxes and lists. Reply text is not available over MCP; read replies in Oppora.

get_dashboard_analyticsOrg-wide totals: lists, leads, sends, opens, replies, bounces, LinkedIn connections, workflows, top campaigns.
Free
get_campaign_analyticsAll-campaign summary or one campaign, with lead status breakdown and follow-up progress.
Free
Arguments
campaign_id?: int, campaign_name?: string (omit both for all campaigns)
get_sequence_campaign_analyticsFull sequence analytics: email, LinkedIn, calls, CRM pushes and a per-step table.
Free
Arguments
campaign_id?: int, campaign_name?: string
Returns

TOON { campaign, audience funnel, rates, email, linkedin, calls, crm, reply_outcomes, steps, tracking_setup }

get_sequence_recipientsPer-lead sequence status: current step, activity on each channel, reply status.
Free
Arguments
campaign_id?: int, campaign_name?: string, filter?: string (all, replied, in_progress, queued, completed, failed, paused, waiting_task, email_sent, opened, clicked, bounced, unsubscribed, invite_sent, invite_pending, connected, li_message, inmail_sent, skipped, unreachable, removed), search?: string, limit?: int (1-100, default 25)
get_pipeline_funnelOutbound funnel from contacted to opportunity: open, in progress, won, lost.
Free
get_email_account_healthPer-mailbox connection, warmup status, daily usage and pause state.
Free
get_list_analyticsList quality: lead and company counts, email coverage, verified and risk rates.
Free
Arguments
list_id?: int, list_name?: string

Research / web

5

Web, news, maps and semantic search. web_search costs 1 data credit and web_scrape 5 data credits per call that returns results; the other three cost 1 AI credit per call that returns results.

web_searchGoogle web search for prospect, competitor and market research: titles, snippets, links and the knowledge graph.
1 data credit per call
web_scrapeFetch a URL and return its readable text.
5 data credits per call
smart_web_searchSemantic web search for niche companies, people, news and market reports.
1 AI credit per call
local_business_searchGoogle Maps businesses with address, phone, website and rating.
1 AI credit per call
news_searchRecent news articles, useful for funding, launch and hiring signals.
1 AI credit per call

End-to-end example

One conversation can take a list from search to a launched sequence. Background steps return a task_id; the assistant polls get_task_status or list_my_tasks.

  1. 1
    get_available_industries + get_available_countries(resolve exact filter values)
  2. 2
    discover_companies(industry=[…], hq_country=[…])(rows carry Oppora ids)
  3. 3
    my_lists or create_list(pick the target list)
  4. 4
    add_companies_to_list(list_id, company_ids=[…])(1 data credit per new company)
  5. 5
    enrich_companies(list_id)(background task)
  6. 6
    find_leads_for_companies(list_id, departments=[…], management_level=[…], title=[…])(background task)
  7. 7
    score_leads(list_id, service_type, leads_per_company, …)(Yes or No per lead)
  8. 8
    find_emails_for_leads(list_id)(background task)
  9. 9
    get_task_status(task_id) or list_my_tasks(status="completed")(progress and links)
  10. 10
    create_sequence_campaign + save_sequence + attach_sequence_audience(draft with email and LinkedIn steps)
  11. 11
    get_sequence, then activate_sequence when you say "launch"(sends for real)
For repeatable runs: create_workflow, then add_workflow_node with steps such as company-crawl, lead-search, ai-recommendation, search-contact and generate-campaign, then set_workflow_icp to target your ICP and start_workflow. For outside data (web scrapes, news, pasted lists): create_companies_bulk and create_leads_bulk save up to 100 rows per call.

FAQ

Which MCP clients are supported?

Any client that supports remote MCP servers over Streamable HTTP. Claude (web, desktop and Claude Code) and ChatGPT connect by signing in with your Oppora account. Cursor, VS Code and other clients connect with an API key. Clients that only run local servers can connect through the mcp-remote bridge.

Do I need an API key?

Not for Claude or ChatGPT. Add the connector URL https://api.oppora.ai/mcp/app, sign in with your Oppora account, pick the organization and click Allow. Cursor, VS Code and scripts use an API key with https://api.oppora.ai/mcp.

Does Oppora's MCP server work with ChatGPT?

Yes. Add a custom MCP app with the URL https://api.oppora.ai/mcp/app, choose OAuth as the authentication and sign in with your Oppora account. Which ChatGPT plans offer custom apps depends on OpenAI; workspace admins may need to allow them.

How do I disconnect Claude or ChatGPT?

In Oppora, open Integrations > Oppora API. Each Claude or ChatGPT connection is listed as a connected app; click Disconnect and it loses access at once. You can also remove the connector in Claude or ChatGPT. To use it again, connect and sign in again.

Is the MCP server free?

Yes. Connecting costs nothing and works on every Oppora plan, including the 7-day free trial. Tool calls spend the same credits as the app. Pro allows 300 a minute and Max 1,500.

How much does each tool cost?

Most tools are free, including lists, filters, templates, campaigns, sequences, workflows, ICP and analytics. Paid tools: web_search 1 data credit and web_scrape 5 data credits per call with results; enrich_company 10 data credits per company found (20 when its tech stack is returned); enrich_person 5 data credits per match; live_people_search 6 data credits per page of 25; linkedin_profile_activity, linkedin_company_posts, linkedin_search_posts and linkedin_post_engagement 1 data credit per page of up to 100; linkedin_get_post 3 data credits; discover_companies costs 5 data credits and discover_people 10 per call with results; search_jobs 1 data credit per page; add_companies_to_list and add_leads_to_list 1 data credit per new item; email_search 1 data credit per email found; email_verifier 1 data credit (standard) or 3 (advanced); phone_search 1 phone credit per number found; find_leads_for_companies 1, 3 or 5 data credits per lead saved; find_emails_for_leads data credits per email found; enrich_companies 1 data credit per field filled; score_leads 1 AI credit per lead; smart_web_search, local_business_search and news_search 1 AI credit per call with results. Misses and empty results are free.

Can the AI send emails without asking?

No tool sends on the AI's own initiative. Outreach starts only through a launch call: start_campaign, or a status change to active through update_campaign_status or create_or_update_campaign, for email campaigns, and activate_sequence or resume_sequence for sequences. Each runs a readiness check and refuses if setup is incomplete. The sequence launch tools tell the assistant to call them only when you explicitly ask, and your request is the go-ahead; Oppora does not add a second confirmation. start_campaign, update_campaign_status, activate_sequence and resume_sequence carry the MCP destructive hint, which many clients use to ask before running a tool. create_sequence_campaign and the editing tools only build drafts.

Do MCP and the REST API use the same key?

Yes, for clients that use a key. One opp_live_ key works for both. Create keys in the app under Integrations > Oppora API > New key; the secret is shown once. Claude and ChatGPT do not need a key.

How does authentication work?

Claude and ChatGPT use OAuth: you sign in to Oppora, choose an organization and approve the connection, and the assistant never sees your password or a key. Other clients send an Oppora API key as Authorization: Bearer YOUR_KEY or X-API-Key: YOUR_KEY. Each connection or key belongs to one user in one organization.

What format do tool responses use?

TOON, a compact tabular text format, for every tool except discover_companies and discover_people, which return JSON because their rows carry nested fields. TOON spends fewer tokens than JSON on lists of uniform records.

How do I check that my key works?

Run MCP Inspector with npx @modelcontextprotocol/inspector, connect to https://api.oppora.ai/mcp over Streamable HTTP with the header Authorization: Bearer YOUR_KEY and list the tools. You should see 81.