
by Orthogonal
Give AI agents instant B2B lead discovery via a single API key.
Find Leads suits sales engineers and founders who want automated B2B prospecting without setting up 27+ separate data-provider accounts and API keys. Give your agent a target persona and it returns matching contacts through Orthogonal's single-key catalog, ready for outreach or CRM import. Fits agent pipelines better than enterprise teams already locked into Apollo.io contracts.
One API key replacing 30+ B2B data provider accounts, with the cost of each call returned in the API response for real-time budget control.
The unified API key model reduces switching costs for developers already using Orthogonal, but data providers like Apollo.io could build their own SKILL.md wrappers, removing the aggregation advantage.
Main trade-off
Pay-per-call pricing becomes more expensive than direct provider subscriptions at high query volumes, and data accuracy depends on third-party providers Orthogonal does not control.
Pay-per-call credits, no monthly subscription required. A starter credit balance of $10 is granted automatically on signup, no card needed. Credits draw down per request at whatever rate the underlying data provider charges. An enterprise tier is available for high-volume workloads.
orth skills add orthogonal/find-leads
Installed via the Orthogonal CLI (orth skills add orthogonal/find-leads), which copies a SKILL.md file into the agent's skills directory. Alternatively, add https://mcp.orth.sh as an MCP server so the agent calls it as an MCP tool. The agent then calls orth run orthogonal/find-leads /search or /enrich with a JSON body containing search criteria.
The permissions this connector asks for, as its own documentation states them.
3 actions an agent can call. Open one for its parameters and an example.
Example call
orth run orthogonal/find-leads /search -d '{
"job_title": "VP Engineering",
"industry": "SaaS",
"company_size": "51-200",
"location": "United States",
"limit": 25
}'Parameters
| Parameter | Type | Needed | Description |
|---|---|---|---|
| job_title | string | Optional | Target job title(s) to filter by, e.g. 'CTO', 'VP Sales', 'Head of Marketing'. |
| industry | string | Optional | Industry vertical, e.g. 'SaaS', 'FinTech', 'Healthcare', 'E-commerce'. |
| company_size | string | Optional | Employee headcount range, e.g. '1-10', '11-50', '51-200', '201-1000', '1000+'. |
| location | string | Optional | Geographic target such as a city, state, or country, e.g. 'San Francisco', 'United States', 'Europe'. |
| keywords | string | Optional | Additional freetext keywords to narrow results, e.g. 'machine learning' or 'Series B'. |
| limit | numberdefault 10 | Optional | Maximum number of lead records to return. |
Returns
| Field | Type | Description |
|---|---|---|
| leads | array | List of lead objects, each with name, company, job_title, linkedin_url, and (when available) email. |
| total_found | number | Total number of matching records in the underlying data source before the limit is applied. |
| cost | number | Credits deducted for this call, returned in every Orthogonal API response. |
Example call
orth run orthogonal/find-leads /enrich -d '{
"name": "Jane Smith",
"company": "Acme Corp",
"domain": "acme.com",
"linkedin_url": "https://linkedin.com/in/janesmith"
}'Parameters
| Parameter | Type | Needed | Description |
|---|---|---|---|
| name | string | Required | Full name of the person to enrich. |
| company | string | Optional | Company name to narrow the person match when the name is common. |
| domain | string | Optional | Company domain for high-precision email matching, e.g. 'acme.com'. |
| linkedin_url | string | Optional | LinkedIn profile URL for the strongest possible person-matching confidence. |
Returns
| Field | Type | Description |
|---|---|---|
| string | Verified work email address. | |
| phone | string | Direct phone number, if available from the data provider. |
| title | string | Current job title as reported by the data source. |
| company_data | object | Enriched company record including headcount, revenue range, industry, and website. |
| social_profiles | array | List of social profile URLs such as LinkedIn and Twitter. |
| cost | number | Credits deducted for this enrichment call. |
Example call
orth run orthogonal/find-leads /export -d '{
"leads": [
{ "name": "Priya Shah", "company": "Globex" },
{ "name": "Bob Lee", "company": "Initech" }
],
"format": "json"
}'Parameters
| Parameter | Type | Needed | Description |
|---|---|---|---|
| leads | array | Required | Array of lead objects (from a prior Find Leads call) to export and normalize. |
| format | stringdefault json | Optional | Output format: 'json' (default) or 'csv'. |
| fields | array | Optional | Subset of fields to include in the export, e.g. ['name', 'email', 'company', 'job_title']. |
Returns
| Field | Type | Description |
|---|---|---|
| data | string | Exported lead records as a JSON string or CSV string depending on the format parameter. |
| record_count | number | Number of records included in the export. |
Orthogonal Find Leads is a SKILL.md-format agent skill from Orthogonal, a Y Combinator W26 company that launched in January 2026. It lets AI agents discover B2B prospects by searching job title, industry, company size, and location across a catalog of data providers including Apollo.io and People Data Labs, returning a ranked contact list through one API key.
Single key covers 30+ data providers
One ORTHOGONAL_API_KEY replaces Apollo.io, People Data Labs, ContactOut, and 27+ other B2B data subscriptions, eliminating per-provider onboarding.
SKILL.md format works in every major agent
The skill installs as a plain-text SKILL.md file readable by Claude Code, Cursor, Windsurf, GitHub Copilot, and Codex CLI with no binary dependencies.
MCP server for zero-install usage
Point any MCP-compatible client at https://mcp.orth.sh and the Find Leads capability is available immediately without running a local CLI.
Pay-per-call with built-in cost visibility
Every API response includes the credit cost of that call, so agents can track spend in real time without a separate billing dashboard.
Automatic rate limiting and retries
Orthogonal handles provider-level rate limits and retry logic transparently, so agents never need to implement backoff logic themselves.
Built-in enrichment and export actions
Beyond search, the skill's Enrich Lead action appends a verified email, phone number, and company data to a known contact, and Export Lead List normalizes a batch of results into JSON or CSV.
Target persona list building
A sales agent searches for 50 VP Engineering contacts at Series B SaaS companies in North America, returning a structured list ready for CRM import in under 30 seconds.
Account-based marketing enrichment
A marketing agent queries a list of target accounts by company size and industry, then passes the results to a lead-enrichment step that appends verified emails before launching an email sequence.
Investor outreach prospecting
A fundraising agent finds active seed-stage investors who have backed B2B SaaS companies in the last 12 months, using Orthogonal's data to build a cold outreach list without manual research.
Competitive hiring intelligence
A talent agent locates engineers with specific skills at target competitor companies, giving a recruiter a vetted shortlist without browsing LinkedIn manually.
The nearest AI agent skills on Hokai for the same job.
| Skill | Works with | Connects by | Checked |
|---|---|---|---|
| Claude · Codex · Cursor · Windsurf +1 | MCP | 19 Aug | |
| -- | MCP | 17 Nov | |
| Claude · Cursor | MCP | 26 Oct | |
| Claude · Cursor | REST API | 11 Oct |
There's no monthly subscription: Orthogonal deducts credits from your balance every time the skill runs a search or enrichment call, and the exact cost is returned with that same response. Pricing is pay-per-call and set by whichever underlying data provider serviced the request, so a call routed to a premium provider costs more than one routed to a cheaper one. New accounts also start with a small free credit balance (see the next question for the exact amount), and heavy users can move to an enterprise tier with custom credit packages.
A starter balance of $10 in credit lands automatically when you sign up, no credit card required, enough for dozens to hundreds of lookups depending on which data provider answers the request. There's no unlimited free tier: once that balance runs out, every further call draws down real credits at the provider's rate. Teams can top up manually, or opt into automatic top-up, which is currently in beta.
Apollo.io's direct API and People Data Labs both offer B2B contact search with similar coverage, but each requires its own account, API key, and billing relationship instead of Orthogonal's single-key catalog. Clay orchestrates several data providers inside a no-code workflow builder and is priced per seat rather than per call, which suits teams that want a visual pipeline over an agent-native one. Hunter.io focuses on email discovery specifically and is a lighter pick when a verified email is the only data point needed.
Apollo.io's own agent skill plugs directly into its dedicated contact database with native intent scoring and CRM sync, but requires an active Apollo.io subscription (roughly $49 per seat per month) and a separate API key. Orthogonal queries Apollo as one source among many in its catalog, so you pay per call instead of per seat, but lose Apollo-specific extras like intent signals and email sequencing. Teams with an existing Apollo.io contract get more depth and cheaper per-record cost by going direct; teams without one get a faster, cheaper start through Orthogonal.
Run npm install -g @orth/cli to get the Orthogonal command-line tool (Node.js 18 or higher), then run orth skills add orthogonal/find-leads to copy the skill into your agent's directory and set ORTHOGONAL_API_KEY as an environment variable. Skip the CLI entirely by adding https://mcp.orth.sh as an MCP server in your agent's config instead. Once installed, call orth run orthogonal/find-leads /search with a JSON body describing who you're searching for (role, industry, headcount, and geography) to get back a ranked contact list within seconds.
Orthogonal Find Leads is an agent skill from Orthogonal, a San Francisco unified API gateway that launched in January 2026 out of Y Combinator's W26 batch. Its founders, Christian Pickett and Bera Sogut, previously worked on payments at Coinbase and billing at Vercel, and on reCAPTCHA, Maps, and Amazon Robotics, respectively.
The skill teaches any SKILL.md-compatible agent how to discover B2B prospects: submit search criteria such as job title, company size, industry, and location, and it queries Orthogonal's catalog of data providers, including Apollo.io, People Data Labs, ContactOut, and Ocean.io, for a ranked list of matching contacts ready for outreach or CRM import. Agents call the Orthogonal REST API or the hosted MCP server at mcp.orth.sh, authenticated with a single ORTHOGONAL_API_KEY that covers every provider in the catalog, so there is no per-provider key management or separate billing relationship to set up.
Find Leads works best inside automated GTM pipelines: a sales agent can identify target CFOs at Series B fintech companies, hand the list to a lead-enrichment step for verified emails, then pass the enriched contacts to an outreach skill. The skill is compatible with Claude Code, Claude Desktop (via MCP), Cursor, Windsurf, GitHub Copilot, Codex CLI, and OpenClaw.
Protocol, authentication, permissions and the call contracts on this page are read from Orthogonal's own documentation. The verdict is Hokai's reading of them, not the maintainer's claim.