Integration

Clay integration

Clay is a go-to-market data and enrichment platform. With this API an agent can search Clay's database of people and companies with Clay's advanced query language and page through the results, run a Clay routine (a Clay-managed or custom enrichment function, which spends the workspace's Clay credits) on up to 100 items and read its results, query known Clay tables (Enterprise plans), search Workflow runs (beta), read the workspace's credit balances and the key's user and workspace, and queue privacy deletions of people from Audiences and follow them. Search results count against plan quotas: Free 50 results per request and 100 per 30 days; paid plans up to 500 per request. Left out: the deprecated filters-mode search routes, and starting a batch routine run, which needs a JSONL file uploaded to a presigned URL that these tools cannot reach (get_batch_run_results still reads a batch started with the Clay CLI or another API client). The API is rate limited per workspace (HTTP 429 with Retry-After). Authentication: a personal Clay API key created under API and CLI, API keys in the Clay app, sent in the clay-api-key header; it acts with that user's workspace access.

Connect Clay How integrations work

Ask it in chat

Say what should happen.
The workflow builds itself.

When this happens, do that in Clay.Built from the 13 Clay tools below, on the connection you add once.

What you can do with Clay

Each of these is one operation on one endpoint, available both as a workflow step and as a tool an AI agent can call on its own. Nothing bundles several operations into one action, so what a run did stays readable afterwards.

Most endpoints run on the free plan. A few, mainly the ones that publish on your behalf, need a paid plan.

Routine run, 2 operations

get_batch_run_resultsGet batch run resultsGETReturns a batch routine run started with the Clay CLI or another API client: in_progress with total and finished, then complete with result_url (a link to download the results), or validation_failed / processing_failed with error.
get_routine_run_resultsGet routine run resultsGETReturns a routine run from run_routine: status in_progress (total, finished: call again later) or complete with data, one item per input id (status complete or failed, result or error). Page with cursor.

Audience person, 1 operation

privacy_delete_peoplePrivacy delete peoplePOSTDELETES the matching people from the workspace's Audiences (privacy requests); cannot be undone. dry_run true only counts matches. A real run answers job_id: follow it with get_privacy_deletion_status.

Credit balance, 1 operation

get_credit_balanceGet credit balanceGETReturns the workspace's remaining Clay credits: balance (spendable data credits, rollover included) and action_execution_balance (remaining actions; absent on legacy or no billing plan). Reading is free.

Privacy deletion, 1 operation

get_privacy_deletion_statusGet privacy deletion statusGETReturns a privacy deletion job: status in_progress (call again later), complete, or failed (cleanup may be partial), with total people selected and finished. Kept up to five days after it ends.

Routine, 1 operation

run_routineRun routinePOSTStarts a Clay routine (an enrichment function) on 1 to 100 items; it SPENDS the workspace's Clay credits. Answers routine_run_id at once: read the results with get_routine_run_results.

Search, 1 operation

create_searchCreate searchPOSTCreates a search of Clay's people or companies database from an advanced query (read get_search_query_reference first) and answers search_id and source_type. Fetch results with get_next_search_results.

Search reference, 1 operation

get_search_query_referenceGet search query referenceGETReturns the reference document (Markdown, in reference) for Clay's advanced search query language: the queryable fields of people and companies and the grammar. Read it before create_search.

Search result, 1 operation

get_next_search_resultsGet next search resultsPOSTReturns the next page of a search's people or companies and advances it: call again while has_more is true. Each call counts against the plan's result quota (period_quota).

Table, 1 operation

query_tablesQuery tablesPOSTQueries known Clay tables (Enterprise plans): pick tables by ID (from the table URL, t_...), select fields, filter, join, group, order. Answers data rows of cells (status, value) and a cursor for the next page.

User, 1 operation

get_meGet meGETReturns the Clay user the API key belongs to (id, name, email) and its workspace (id, name). Useful to check the connection.

Workflow run, 1 operation

query_workflow_runsQuery workflow runsPOSTSearches Clay Workflow runs (beta) with Clay's run query language (see get_workflow_runs_query_reference): a select query answers matching runs newest first with a cursor, a count query answers count.

Workflow run reference, 1 operation

get_workflow_runs_query_referenceGet workflow runs query referenceGETReturns the reference document (Markdown, in reference) for Clay's Workflow run query language: queryable run fields, node predicates, content search and grammar. Read it before query_workflow_runs.

Use Clay from Claude Code, Cursor or Codex

Connect Clay in LiveContext, create an API key in Settings, then add the LiveContext MCP server to your client. Its Clay operations become tools the client can call, next to every other integration you connected.

Claude Code

claude mcp add --transport http livecontext https://livecontext.ai/mcp --header "X-API-Key: YOUR_API_KEY"

Codex (~/.codex/config.toml)

[mcp_servers.livecontext]
url = "https://livecontext.ai/mcp"

[mcp_servers.livecontext.http_headers]
"X-API-Key" = "YOUR_API_KEY"

Cursor, Claude Desktop and other JSON configs

{
  "mcpServers": {
    "livecontext": {
      "type": "http",
      "url": "https://livecontext.ai/mcp",
      "headers": {
        "X-API-Key": "YOUR_API_KEY"
      }
    }
  }
}

Set up the MCP server step by step

How to connect Clay

Create your workspace
Sign up free. Nothing to install.
Connect Clay once
Add your Clay connection in Settings. Every tool from this integration reuses it, and secrets are encrypted at rest and visible only to you.
Describe the job in chat
The workflow builds itself in front of you and picks the endpoints it needs. Drag the steps yourself on the canvas whenever you would rather be explicit.