Skip to content

Update to catchmeup to support user list and pronouns #482

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Oct 1, 2025
Merged

Update to catchmeup to support user list and pronouns #482

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
115 changes: 69 additions & 46 deletions Parsers/Summarize the thread so far.js
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,92 +5,115 @@ flags:gmi
*/

(function(current) {
var si = new Slacker();
var thread_ts = current.getValue('thread_ts');

// Commented out code block
// if (!thread_ts) {
// message = 'This command is only usable in a thread.';
// si.send_chat(current, message, false);
// return null;
// }

function findWordAfterDashDash(str) {
var match = str.match(/--\s*(.+)/);
return match ? " in the style of " + match[1] : "";
let match = str.match(/--\s*(.+)/);
return match ? ` in the style of ${match[1]}` : '';
}
var style = findWordAfterDashDash(current.text);

var chats = [];
var nda = false;
var chatGr = new GlideRecord('x_snc_pointsthing_chat');
const si = new x_snc_slackerbot.Slacker();
const thread_ts = current.getValue('thread_ts');
const style = findWordAfterDashDash(current.text);
const chats = [];
const userPronouns = [];
const users = [];
let nda = false;
let encodedQuery = '';
if (thread_ts) {
chatGr.addEncodedQuery('thread_ts=' + thread_ts + '^ORts=' + thread_ts + '^textNOT LIKE!catchmeup^textNOT LIKE!summary^textNOT LIKE!catchup');
encodedQuery = `thread_ts=${thread_ts}^ORts=${thread_ts}^textNOT LIKE!catchmeup^textNOT LIKE!summary^textNOT LIKE!catchup`;
} else {
chatGr.addEncodedQuery('channel=' + current.getValue('channel') + '^sys_created_onRELATIVEGT@minute@ago@60^thread_ts=NULL^textNOT LIKE!catchmeup^textNOT LIKE!summary^textNOT LIKE!catchup');
encodedQuery = `channel=${current.getValue('channel')}^sys_created_onRELATIVEGT@minute@ago@60^thread_ts=NULL^textNOT LIKE!catchmeup^textNOT LIKE!summary^textNOT LIKE!catchup`;
}


const chatGr = new GlideRecord('x_snc_slackerbot_chat');
chatGr.addEncodedQuery(encodedQuery);
chatGr.orderBy('sys_created_on');
chatGr.query();

while (chatGr.next()) {
var chat = chatGr.getDisplayValue('user') + ': ' + chatGr.getValue('text');
const chat = `${chatGr.getDisplayValue('user')}: ${chatGr.getValue('text')}`;
if (chat.indexOf('!nda') > -1 || chat.indexOf('!confidential') > -1) {
nda = true;
break;
}
chats.push(chat);
}

if (nda) {
if (thread_ts) {
new x_snc_slackerbot.Slacker().send_chat(current, "This thread has been marked confidential and cannot be summarized.", false);
si.send_chat(current, 'This thread has been marked confidential and cannot be summarized.', false);
} else {
new x_snc_slackerbot.Slacker().send_chat(current, "This recent conversation has been marked confidential and cannot be summarized.", false);
si.send_chat(current, 'This recent conversation has been marked confidential and cannot be summarized.', false);
}
return;
}

var prompt = current.text.replace(/!chatgpt/gmi, "").trim().substring(0, 1000);
var chatReq = new sn_ws.RESTMessageV2();
const userGa = new GlideAggregate('x_snc_slackerbot_chat');
userGa.addEncodedQuery(encodedQuery);
userGa.addAggregate('GROUP_CONCAT_DISTINCT','user.user_id');
userGa.query();

while(userGa.next()){
const userId = userGa.getAggregate('GROUP_CONCAT_DISTINCT','user.user_id');

const rm = new sn_ws.RESTMessageV2();
rm.setHttpMethod('GET');
rm.setLogLevel('all');
rm.setRequestHeader('Content-Type', 'application/x-www-form-urlencoded');
rm.setRequestHeader('authorization', `Bearer ${gs.urlEncode(gs.getProperty('x_snc_slackerbot.SlackerBot.token'))}`);
rm.setEndpoint(`https://slack.com/api/users.info?user=${userId}`);
const response = rm.execute();
const responseBody = JSON.parse(response.getBody());
const pronouns = responseBody.user.profile.pronouns ?? 'they/them';
const name = responseBody.user.real_name ?? 'Unknown';
userPronouns.push(`${name}: ${pronouns}`);
users.push(name);
}

const chatReq = new sn_ws.RESTMessageV2();
chatReq.setEndpoint('https://api.openai.com/v1/chat/completions');
chatReq.setHttpMethod("POST");
chatReq.setRequestHeader("Authorization", "Bearer " + gs.getProperty("openai.key"));
chatReq.setRequestHeader('Content-Type', "application/json");
chatReq.setRequestHeader('User-Agent', "ServiceNow");
chatReq.setRequestHeader("Accept", "*/*");
var body = {
"model": "gpt-4o",
"messages": [{
"role": "system",
"content": `
chatReq.setHttpMethod('POST');
chatReq.setRequestHeader('Authorization', `Bearer ${gs.getProperty('openai.key')}`);
chatReq.setRequestHeader('Content-Type', 'application/json');
chatReq.setRequestHeader('User-Agent', 'ServiceNow');
chatReq.setRequestHeader('Accept', '*/*');
const body = {
'model': 'gpt-4o',
'messages': [{
'role': 'system',
'content': `
You are a Slack bot that enhances the formatting of messages to make them more engaging and visually appealing.
Utilize Slack's markdown language mrkdwn for various formatting elements.
Follow these instructions:
1. Enclose important words or phrases with *asterisks* for bold emphasis.
2. Enclose code and numbers and percentages using backticks, like \`this\`.
3. Use emojis when necessary to add expressiveness.
4. Organize text with numbered or bullet lists, using "-" for bullet points.
4. Organize text with numbered or bullet lists, using '-' for bullet points.
5. Combine bold and lists as needed: *Bold text*: normal text.
6. Italicize words like _this_.
7. Use blockquotes with ">" for quotes.
7. Use blockquotes with '>' for quotes.
8. For URLs, use <http://example.com|Clickable link>.
9. Keep user (@user) and channel (#channel) tags unchanged.
Keep close to the original message tone and formatting.`
Keep close to the original message tone and formatting.

Use the below mapping between user name and pronouns to ensure you use the correct pronoun:
${userPronouns.join('\n')}`
},
{
"role": "user",
"content": "summarize the following conversation" + style + ". You cannot ask for follow-up responses. Ignore the user named Slackbot.\n\n" + chats.join("\n")
'role': 'user',
'content': `summarize the following conversation${style}. You cannot ask for follow-up responses. Ignore the user named Slackbot.\n\n${chats.join('\n')}`
}
],
// "max_tokens": 250
// 'max_tokens': 250
};
chatReq.setRequestBody(JSON.stringify(body));
var chatResponse = chatReq.execute();
const chatResponse = chatReq.execute();
gs.info(chatResponse.getBody());
var chatResponseBody = JSON.parse(chatResponse.getBody());
const chatResponseBody = JSON.parse(chatResponse.getBody());

var show_tokens = false;
var token_cost = show_tokens ? "> tokens: " + chatResponseBody.usage.total_tokens + " ($" + (parseInt(chatResponseBody.usage.total_tokens) * 0.000002).toFixed(6) + ")\n" : "";
const show_tokens = false;
const token_cost = show_tokens ? `> tokens: ${chatResponseBody.usage.total_tokens} ($${(parseInt(chatResponseBody.usage.total_tokens) * 0.000002).toFixed(6)})\n` : '';

var intro = thread_ts ? "> This thread so far:\n" : "> Last 60 minutes summarized:\n";
new x_snc_slackerbot.Slacker().send_chat(current, intro + token_cost + "\n" + chatResponseBody.choices[0].message.content, false);
const intro = thread_ts ? '> This thread so far:\n' : '> Last 60 minutes summarized:\n';
si.send_chat(current, `${intro}\nInvolved users:\n${users.join('\n')}\n${token_cost}\n${chatResponseBody.choices[0].message.content}`, false);
})(current);
Loading