heyitsaamir

Agent Provisioning for Teams

Creation

Note

This guide mainly covers Provisioning, not creation of the agent. But here is a simplified version of that...

If you use the sdk's CLI, you can have an app set up in just a few commands. Here's the command for python for example:

npx @microsoft/teams.cli new python my-super-agent -t echo

heyitsaamir

README.md

I use these snippets in my Teams AI Library projects (or other projects).

heyitsaamir

Teams App Development Workflow (Without ATK Extension)

This guide explains how to run Teams apps using the Teams AI Library without relying on the ATK VS Code extension, using custom shell functions for a more controlled development experience.

Overview

Instead of using the ATK VS Code extension (which recreates devtunnels on every launch), this approach uses a long-living devtunnel and custom functions to streamline the development process.

Prerequisites

heyitsaamir

from datetime import datetime, timedelta
import re
from typing import List, Union, Tuple
from autogen.agentchat.contrib.vectordb.base import QueryResults, VectorDB, Document
from autogen.agentchat.contrib.retrieve_user_proxy_agent import RetrieveUserProxyAgent
from autogen.agentchat.contrib.retrieve_assistant_agent import RetrieveAssistantAgent
from autogen import AssistantAgent, ConversableAgent
from azure.core.credentials import AzureKeyCredential
from azure.search.documents import SearchClient
from azure.core.exceptions import ResourceNotFoundError

heyitsaamir

Convert To gif

If you're on a mac, and you use cmd+shift+5 to record screens to put in your PR, here's a little snippet i wrote to convert that into a gif:

convertToGif () {
  if [ -z "$1" ]; then
    mov_filename=$(ls -t ~/Desktop/*.mov | head -n 1)
    if [ -z "$mov_filename" ]; then
      echo "No MOV files found on the desktop."

heyitsaamir

/* globals jQuery, $, waitForKeyElements */

// ==UserScript==
// @name         Github Title Selector
// @namespace    http://tampermonkey.net/
// @version      0.1
// @description  try to take over the world!
// @author       Aamir Jawaid
// @match        https://github.com/*/pull/*
// @icon         https://www.google.com/s2/favicons?domain=tampermonkey.net

heyitsaamir

import React, { useContext, useRef } from 'react';

import { isEqual } from 'lodash';
import { Range } from 'slate';
import { ReactEditor, useSlate } from 'slate-react';

export const CurrentSelectionContext = React.createContext<{
  current: Range | null;
}>({ current: null });

heyitsaamir

with prices as (
SELECT entity_id, id, created_at
      FROM src.prices p
      WHERE p.org_type = 'CARRIER'
      AND p.source_type = 'MODEL'
      AND p.reason_code != 'DEDICATION_PRICE_REGEN'
      QUALIFY ROW_NUMBER() OVER (PARTITION BY p.entity_id ORDER BY p.created_at desc) = 1
      )
, total_prices as (
select p.entity_id, sum(pli.cents) total, LISTAGG(pli.type, ', ') AS cost_type

heyitsaamir

 
with historical_shipment_matching as 
    (
    SELECT -- Use a distinct to avoid duplicate entries coming from lane mapping table
     DISTINCT s.id AS shipment_id
            , s.created_at as matched_at
            , m.contract_id
            , coalesce(m.rate_override, cc.contract_rate) AS contracted_rate_dollars
            , s.pickup_driver_end_time

heyitsaamir

with qualified as (
select *, get_path(matching_criteria, 'jobSummary.shipmentDigests[0].classification') classification from src.dedication_event_logs del
where event_type = 'MATCHING'
and match_type = 'QUALIFIED'
and matching_contracts != '[]'
)
, drop_drops as (
select s.id, 'DROP_DROP' classification from bi.SHIPMENTS s
inner join src.shipment_stops ss_pick on ss_pick.id = s.pickup_stop_id
inner join src.shipment_stops ss_drop on ss_drop.id = s.last_dropoff_stop_id