agentleFS
Talkif-aiTalkif-ai/skills/skills/quickstart/SKILL.md

quickstart

Skill0 starsChanged 2 months agoMITView on GitHub

End-to-end orchestration for a first Talkif integration — API key → flow from a template → phone number → connect → test call → read results — naming which skill to load at each step. Use as the entry point when asked to "set up Talkif", "build a phone agent", or "make my first call" from code.

  • Reads credentials

9 other projects use a version of this.

Adopted by

Most recently taken up by

The file

Files
SKILL.md

Talkif quickstart (orchestration)

Load the named skill for detail at each step; this skill only sequences them.

  1. Authenticate — setup-api-key. Result: client with a key that has calls:*, flows:*, flow_templates:*, phones:*, contacts:*. Verify with GET /calls.
  2. Create a flow — create-flow. Fastest: flowTemplates.instantiateTemplate("customer-support", { name }), then flows.updateFlow(id, { definition }) to set systemPrompt/agentPrompt (write-prompts). Attach any backend calls (create-flow-function).
  3. Validate and publish — flows.validateFlow(id) → fix → flows.publishFlow(id, { notes }). Only a published flow takes calls.
  4. Get a line — manage-phone-numbers: system provider → search by API → the account Owner buys it in the dashboard (an API key can't buy: 403 api_key_not_allowed) → find it with phoneNumbers.listPhoneNumbers. Until the account's first payment, the welcome credit can't pay for a number: the Owner tops up first.
  5. Connect inbound — phoneNumbers.connectFlow(numberId, { flowId }). Dialing the number now reaches the agent (handle-inbound-calls).
  6. Test — test-a-flow: place a real call to the user's own phone with calls.makeCall, or run the draft in the browser with the authenticated SDK. Treat 202 as success (make-call).
  7. Read results — read-call-results: getCallDetails, getCallTranscript, getBillingCallCostBreakdown. For live updates, realtime-events.
  8. Scale — a list → create-campaign (+ manage-contacts, do-not-call); one person on a cadence → create-schedule; a web button → web-calls; spend alerts → read-billing. Anything failing → troubleshoot-calls.

Minimal working script (TypeScript):

import { TalkifClient } from "@talkif/sdk";
const client = new TalkifClient({ token: process.env.TALKIF_API_KEY });

const flow = await client.flowTemplates.instantiateTemplate("customer-support", { name: "Support line" });
await client.flows.publishFlow(flow.id, { notes: "initial" });

// Buying a number is dashboard-only (the account Owner); use one the account already has.
const { phoneNumbers } = await client.phoneNumbers.listPhoneNumbers({ limit: 100 });
const number = phoneNumbers.find(n => n.phoneNumber === "+1YOURTALKIFNUMBER")!;
await client.phoneNumbers.connectFlow(number.id, { flowId: flow.id });

const result = await client.calls.makeCall({ flowId: flow.id, fromNumber: number.phoneNumber, providerId: number.providerId, toNumber: "+1YOURPHONE" });
console.log(result.type === "initiated" ? result.callId : `queued #${result.position}`);

Docs: https://docs.talkif.ai/getting-started/quickstart

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.