Agent-callable Perplexity tools — ask web-grounded questions with citations, run and poll deep research, list models, and search the web. Use when the user mentions Perplexity or wants a cited web answer or search, even if they don't name Perplexity.
Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.
Une commande directe contourne le prompt de vérification. Examinez la source avant de l'exécuter.
Agent-callable Perplexity tools — ask web-grounded questions with citations, run and poll deep research, list models, and search the web. Use when the user mentions Perplexity or wants a cited web answer or search, even if they don't name Perplexity.
license
Elastic-2.0
compatibility
Run `npm install --omit=dev` in this directory, then `node cli.js`. The TypeScript source needs Node.js 22.18+; on older Node, run `cli.js` for build-it-yourself / prebuilt / alternative-runtime options.
Agent-callable tools for Perplexity: ask a question and get a current, web-grounded answer with citations, poll a long-running research job, discover the available models, and run a ranked web search. Wraps the Perplexity Agent API and Search API (https://api.perplexity.ai).
Independent, unofficial connector for Perplexity. Not affiliated with, endorsed by, or sponsored by Perplexity. "Perplexity" is a trademark of its owner, used only to identify the service this connector works with.
When to use this
Answer a question with up-to-date, cited information gathered from the web.
Run deep research as a background job and poll for the result when it's ready.
Discover which models (including deep-research) are available to use.
Get ranked web search results — titles, URLs, and snippets — for a query.
If the connector has not been installed as a skill yet, install it first with npx skills add zapier/connectors --skill perplexity (or your harness's own skill-install mechanism), then continue here. Installing the skill copies these files, not dependencies. Before running the CLI, a local MCP server, or zapier-sdk auth commands, run npm install --omit=dev here once. Importing the published package as a dependency in your own project instead? That npm install already resolves everything — see references/use-as-sdk.md.
Want the actual repo source instead — to browse references/, run this connector's tests, or hack on it? See README.md for a scoped git clone.
The connector runs on Node.js 22.18+. Pick the reference that matches how you're running it, and load it before doing anything else:
You have...
Load
An MCP-aware client — tools may already be loaded (e.g. mcp__perplexity__<tool>), or you can register a local server yourself (or guide the user to)
No tool access, no terminal, no ability to import this package — you write your own code that calls the Perplexity API directly (e.g. a code-execution sandbox)
All scripts share a single perplexity connection — one Perplexity API key.
Script
Script name
Connections
Description
scripts/createAgent.ts
createAgent
perplexity
Ask a question and get a web-grounded, cited answer. Enable web search, pick a model or preset, request structured output, or run long jobs in the background.
scripts/getAgentResponse.ts
getAgentResponse
perplexity
Retrieve or poll an agent response by id — poll a background / deep-research run until it completes.
scripts/listModels.ts
listModels
perplexity
List the model ids usable with createAgent, including deep-research models.
scripts/search.ts
search
perplexity
Search the web and get ranked results with titles, URLs, and snippets.
Disambiguation & refusals
Unsupported requests — decline, don't fake. This connector answers questions and searches the web; it does not generate images, produce text embeddings, stream partial output, or run the Agent API's other tools (people or finance search, URL fetching, code sandbox). If a user asks for one of these, say plainly that it isn't supported and stop — never substitute a different tool and report success for an action you didn't perform.
Auth
Every shape passes auth as one connection selector, not the secret — a [<resolver>:]<value> string. Every connector accepts zapier:<connection-id> (Zapier-managed auth — routes through Zapier's auth, retries, and governance layer); some also accept one or more direct-token resolvers (naming and count vary per connector) — check this connector's own resolvers rather than assuming. The <resolver>: prefix is optional; a bare value goes to the first resolver that claims it — a UUID-shaped bare value always claims zapier:. Each script declares the connections it needs and the resolvers each accepts. The exact syntax for passing a connection (and how to see this connector's resolver list) differs by shape — see the reference you loaded above.
Checking what's already configured first? Don't dump environment values to do it — env or env | grep <name> prints the value along with the name, leaking a live credential into the transcript if one is set. Check names only (env | cut -d= -f1 | grep -i <name>) or test a known name directly ([ -n "$VAR_NAME" ]).
No connection yet? Pick one — and follow the reference's own flow to obtain it; never just ask the user for a connection id or token as if they already have one memorized:
data — the script's result (the shape its outputSchema declares; see the reference you loaded above for how to inspect a script's exact schema in your shape).
meta.outputDataValidation — what validating data did:
{ skipped: false, droppedPaths: [...], instruction } — validated, but those paths were stripped from data: fields the script returned from the API that the outputSchema doesn't declare. If you need them, re-run with output validation skipped.
{ skipped: true } — validation was bypassed; data is the raw, unchecked script output.
Reading dropped fields / skipOutputDataValidation. To receive the raw, unvalidated result, opt out of output validation (the exact syntax differs by shape — see the reference you loaded above). Input validation is never skipped.
Trimming the result / filterOutputData. To shrink a large result down to the fields you need, pass a jq expression that post-processes data (again, exact syntax per shape). The jq runs against data only, NOT the { data, meta } envelope, so write it rooted at data (run the script's --help — or your shape's equivalent — to see its output schema). The transformed value replaces data, meta is preserved, and the result is NOT re-validated against the output schema.
References
Load the matching reference file before working in that area:
Auth & API keys, error status codes (401/400/429), rate limits & usage tiers, provider/model ids, the Agent request shape (input, model vs preset, web search as a tool, reasoning effort), background runs and terminal statuses, structured-output schema constraints and the first-request schema-prep delay, Search API multi-query/date-filter/max_results rules, and grounding/citations.
Before calling any tool — to get error handling, rate-limit backoff, date-filter formats, model ids, background polling, or structured output right.
