Tavily MCP integration for AI agents.
Connect AI agents to Tavily through 7 structured actions, including get research, create research, and crawl. Review authentication, inputs, workflows, and s…
- ACTION CONTRACTS
- 7
- AUTHENTICATION
- API key
- SOURCE STATUS
- Provider-backed
CATALOG SOURCE REVIEWED AUGUST 23, 2026 / ACTION NAMES AND SCHEMAS DERIVED FROM WORKING MACHINES PROVIDER SOURCE
CAPABILITY PROFILE
What the Tavily integration exposes
The Tavily Agent App exposes 7 provider-backed actions for ai and data work. Its current contract lets an authorized agent get the current status and result for a tavily research task, start an asynchronous tavily research task and return a request id for polling, and crawl a website and extract content from discovered pages with tavily. These operations are called through Working Machines as typed capabilities rather than through browser navigation or copied UI steps.
A connection uses API key. Before execution, the agent can inspect the selected action, its required fields, declared scopes, and expected output contract. Provider credentials remain inside the Working Machines runtime; the calling agent receives the capability and its structured result, not the underlying secret.
For reliable operation, start with the narrowest action that satisfies the task, resolve stable provider identifiers before changing state, and validate the returned object or status after execution. Availability still depends on the connected Tavily account, granted provider permissions, workspace policy, region, plan, and upstream API behavior.
VERIFIED ACTION SAMPLE
Real Tavily capabilities.
Showing 7 of 7 actions. Risk labels are conservative signals based on operation names, not substitutes for provider documentation or runtime policy.
get_researchREADGet the current status and result for a Tavily Research task.
INPUTS: request_id
create_researchWRITEStart an asynchronous Tavily Research task and return a request ID for polling.
INPUTS: input / model / stream / output_schema / citation_format
crawlREVIEWCrawl a website and extract content from discovered pages with Tavily.
INPUTS: url / instructions / max_depth / max_breadth / limit
extractREVIEWExtract structured page content from one or more URLs with Tavily.
INPUTS: urls / query / chunks_per_source / extract_depth / include_images
get_usageREADGet API key and account usage details from Tavily.
mapREVIEWDiscover URLs from a website with Tavily Map.
INPUTS: url / instructions / max_depth / max_breadth / limit
searchREADExecute a Tavily Search query and return ranked source results.
INPUTS: query / search_depth / chunks_per_source / max_results / topic
INPUT CONTRACTS
Know what the action needs before it runs.
get_research
request_idREQUIRED- The unique identifier of the research task.
create_research
inputREQUIRED- The research task or question to investigate.
modelOPTIONAL- Research model to use.
streamOPTIONAL- Must be false or omitted. Tavily SSE streaming is not supported.
output_schemaOPTIONAL- JSON Schema for structured research output.
citation_formatOPTIONAL- The format for citations in the research report.
crawl
urlREQUIRED- The root URL that Tavily should crawl.
instructionsOPTIONAL- Natural-language instructions that guide the crawl.
max_depthOPTIONAL- The maximum crawl depth.
max_breadthOPTIONAL- The maximum number of links to follow per crawl level.
limitOPTIONAL- The maximum number of links Tavily should process.
extract
urlsREQUIRED- The URLs that Tavily should extract content from.
queryOPTIONAL- An optional query used to rerank extracted chunks.
chunks_per_sourceOPTIONAL- The maximum number of chunks to return per source when query is provided.
extract_depthOPTIONAL- Controls whether Tavily uses basic or advanced extraction.
include_imagesOPTIONAL- Whether to include images found on each page.
PROVIDER-SPECIFIC WORKFLOWS
Jobs this Agent App can support
Inspect Get Research
Get the current status and result for a Tavily Research task. Use this as a bounded discovery step, retain the returned identifier, and avoid expanding the read beyond the task's stated scope.
get_researchControl Create Research
Start an asynchronous Tavily Research task and return a request ID for polling. Resolve the target first, present material changes for confirmation, and make retries idempotent where the provider supports it.
create_researchVerify with Get Usage
Get API key and account usage details from Tavily. Compare the returned provider state with the intended outcome and preserve stable IDs or canonical links in the run record.
get_usageSAFETY BOUNDARY
Operate Tavily with explicit limits
- Authorize Tavily with API key and grant only the provider access required by the selected actions.
- 1 action is change-capable by name. Confirm the target identity and material parameters before allowing a write.
- Do not infer permission from catalog visibility. Workspace policy, connection identity, and upstream authorization still govern execution.
- No provider scope string is declared on the sampled actions. Verify the connected account's actual permissions in Tavily rather than assuming unrestricted access.
- After a call, inspect the structured result and execution record before reporting that the Tavily task completed successfully.
CONNECTION MODEL
API key
Tavily API key used with the Authorization: Bearer <apiKey> header.
Working Machines stores provider credentials behind the execution boundary. An agent can use an authorized connection identity, but catalog discovery alone does not reveal OAuth tokens, API keys, or provider secrets.
- READ SIGNALS
- 3
- WRITE SIGNALS
- 1
- HIGH IMPACT
- 0
- REVIEW SIGNALS
- 3
EVIDENCE AND AVAILABILITY
Provider reference
Action names, input fields, authentication types, and counts on this page are generated from the Working Machines provider catalog. Provider behavior, quotas, object semantics, account eligibility, and regional availability remain governed by Tavily.