API

Every tavilymcp.api.* operation is listed below with parameter shapes and return types from the plugin Zod schemas.

New to Corsair? See API access, authentication, and error handling.

Tavily

crawl

tavily.crawl Crawl a site from a start URL with configurable depth and breadth Risk: read

await corsair.tavilymcp.api.tavily.crawl({});

Input

Name	Type	Required	Description
`url`	`string`	Yes	—
`max_depth`	`number`	No	—
`max_breadth`	`number`	No	—
`limit`	`number`	No	—
`instructions`	`string`	No	—
`chunks_per_source`	`number`	No	—
`select_paths`	`string[]`	No	—
`select_domains`	`string[]`	No	—
`exclude_paths`	`string[]`	No	—
`exclude_domains`	`string[]`	No	—
`allow_external`	`boolean`	No	—
`categories`	`Careers \| Blog \| Documentation \| About \| Pricing \| Community \| Developers \| Contact \| Media \| People[]`	No	—
`extract_depth`	`basic \| advanced`	No	—
`format`	`markdown \| text`	No	—
`include_images`	`boolean`	No	—
`include_favicon`	`boolean`	No	—
`timeout`	`number`	No	—
`include_usage`	`boolean`	No	—

Output

Name	Type	Required	Description
`base_url`	`string`	Yes	—
`results`	`object[]`	Yes	—
`failed_results`	`object[]`	No	—
`response_time`	`number`	Yes	—
`usage`	`object`	No	—
`request_id`	`string`	No	—

results full type

{
  url: string,
  raw_content: string,
  images?: string[],
  favicon?: string | null
}[]

failed_results full type

{
  url: string,
  error: string
}[]

usage full type

{
  credits: number
}

extract

tavily.extract Extract page content from one or more URLs as markdown or plain text Risk: read

await corsair.tavilymcp.api.tavily.extract({});

Input

Name	Type	Required	Description
`urls`	`string \| string[]`	Yes	—
`query`	`string`	No	—
`chunks_per_source`	`number`	No	—
`include_images`	`boolean`	No	—
`extract_depth`	`basic \| advanced`	No	—
`format`	`markdown \| text`	No	—
`include_favicon`	`boolean`	No	—
`timeout`	`number`	No	—
`include_usage`	`boolean`	No	—

Output

Name	Type	Required	Description
`results`	`object[]`	Yes	—
`failed_results`	`object[]`	Yes	—
`response_time`	`number`	Yes	—
`usage`	`object`	No	—
`request_id`	`string`	No	—

results full type

{
  url: string,
  raw_content: string,
  title?: string | null,
  images?: string[],
  favicon?: string | null
}[]

failed_results full type

{
  url: string,
  error: string
}[]

usage full type

{
  credits: number
}

map

tavily.map Map a site’s structure from a start URL and return its URLs Risk: read

await corsair.tavilymcp.api.tavily.map({});

Input

Name	Type	Required	Description
`url`	`string`	Yes	—
`max_depth`	`number`	No	—
`max_breadth`	`number`	No	—
`limit`	`number`	No	—
`instructions`	`string`	No	—
`select_paths`	`string[]`	No	—
`select_domains`	`string[]`	No	—
`exclude_paths`	`string[]`	No	—
`exclude_domains`	`string[]`	No	—
`allow_external`	`boolean`	No	—
`categories`	`Careers \| Blog \| Documentation \| About \| Pricing \| Community \| Developers \| Contact \| Media \| People[]`	No	—
`timeout`	`number`	No	—
`include_usage`	`boolean`	No	—

Output

Name	Type	Required	Description
`base_url`	`string`	Yes	—
`results`	`string[]`	Yes	—
`response_time`	`number`	Yes	—
`usage`	`object`	No	—
`request_id`	`string`	No	—

usage full type

{
  credits: number
}

research

tavily.research Run comprehensive multi-source research on a topic and return a cited report Risk: read

await corsair.tavilymcp.api.tavily.research({});

Input

Name	Type	Required	Description
`input`	`string`	Yes	—
`model`	`auto \| mini \| pro`	No	—
`output_schema`	`object`	No	—
`citation_format`	`numbered \| mla \| apa \| chicago`	No	—
`include_domains`	`string[]`	No	—
`exclude_domains`	`string[]`	No	—
`output_length`	`short \| standard \| long`	No	—
`files`	`object[]`	No	—
`max_wait_ms`	`number`	No	—
`poll_interval_ms`	`number`	No	—

output_schema full type

{
}

files full type

{
  name: string,
  data: string,
  type: base64
}[]

Output

Name	Type	Required	Description
`request_id`	`string`	Yes	—
`status`	`pending \| in_progress \| completed \| failed`	Yes	—
`input`	`string`	No	—
`model`	`string`	No	—
`created_at`	`string`	No	—
`content`	`object`	No	—
`sources`	`object[]`	No	—
`response_time`	`number`	No	—
`usage`	`object`	No	—

content full type

string | {
}

sources full type

{
  title?: string | null,
  url: string,
  favicon?: string | null
}[]

usage full type

{
  credits: number
}

search

tavily.search Search the web with Tavily and return ranked snippets with source URLs Risk: read

await corsair.tavilymcp.api.tavily.search({});

Input

Name	Type	Required	Description
`query`	`string`	Yes	—
`search_depth`	`advanced \| basic \| fast \| ultra-fast`	No	—
`chunks_per_source`	`number`	No	—
`max_results`	`number`	No	—
`topic`	`general \| news \| finance`	No	—
`time_range`	`day \| week \| month \| year \| d \| w \| m \| y`	No	—
`start_date`	`string`	No	—
`end_date`	`string`	No	—
`include_answer`	`boolean \| basic \| advanced`	No	—
`include_raw_content`	`boolean \| markdown \| text`	No	—
`include_images`	`boolean`	No	—
`include_image_descriptions`	`boolean`	No	—
`include_favicon`	`boolean`	No	—
`include_domains`	`string[]`	No	—
`exclude_domains`	`string[]`	No	—
`country`	`string`	No	—
`auto_parameters`	`boolean`	No	—
`exact_match`	`boolean`	No	—
`safe_search`	`boolean`	No	—
`include_usage`	`boolean`	No	—

Output

Name	Type	Required	Description
`query`	`string`	Yes	—
`answer`	`string`	No	—
`follow_up_questions`	`string[]`	No	—
`images`	`object[]`	Yes	—
`results`	`object[]`	Yes	—
`response_time`	`number`	Yes	—
`usage`	`object`	No	—
`request_id`	`string`	No	—
`auto_parameters`	`object`	No	—

images full type

{
  url: string,
  description?: string | null
}[]

results full type

{
  title: string,
  url: string,
  content: string,
  score: number,
  raw_content?: string | null,
  favicon?: string | null,
  images?: {
    url: string,
    description?: string | null
  }[] | null,
  id?: string
}[]

usage full type

{
  credits: number
}

auto_parameters full type

{
}

Overview

Plugins

Tavily

crawl

extract

map

research

search

​Tavily

​crawl

​extract

​map

​research

​search

Tavily

crawl

extract

map

research

search