meetings

Name: meetings
Author: butterbase-ai

533stars40forks

How to install meetings?

butterbase-ai/butterbase-skills/meetings

$npx -y skills add butterbase-ai/butterbase-skills --skill meetings

Installs into the current project.

Summary

Use when building features that join, record, or transcribe Zoom/Meet/Teams/Webex calls — meeting bots, call notetakers, sales-call summarizers, interview transcribers. Covers ctx.ai.meetings.start/get/stop/list, webhook setup, and credit metering.

Files · 1

View on GitHub

SKILL.md

1	# Meeting Bots on Butterbase
2
3	Guide for building features that spawn meeting bots via `ctx.ai.meetings`. Covers the four SDK methods, webhook handling, idempotency, and cost-aware design.
4
5	---
6
7	## 1. When to Use
8
9	Reach for this skill when your feature involves any of:
10
11	- Meeting notetaker — bot joins a recurring standup or sales call and stores a transcript.
12	- Sales call enricher — join a customer call, wait for `bot.done`, then run an LLM over the transcript to populate a CRM record.
13	- Interview pipeline — record and transcribe candidate interviews, then score against a rubric.
14	- Voice-of-customer transcripts — capture support or discovery calls; feed chunks into a RAG collection for future search.
15
16	If you need only audio/video file upload without a live meeting join, use `manage_storage` instead.
17
18	---
19
20	## 2. The Four SDK Methods
21
22	\| Method \| What it does \| One-line example \|
23	\|--------\|-------------\|-----------------\|
24	\| `bb.ai.meetings.start(opts)` \| Spawns a bot and returns immediately with `status: "joining"` \| `const { data: bot } = await bb.ai.meetings.start({ meetingUrl, transcript: true })` \|
25	\| `bb.ai.meetings.get(id)` \| Fetches current status + artifact URLs \| `const { data } = await bb.ai.meetings.get(bot.id)` \|
26	\| `bb.ai.meetings.stop(id)` \| Removes the bot from the call early \| `await bb.ai.meetings.stop(bot.id)` \|
27	\| `bb.ai.meetings.list(opts)` \| Lists bots with optional `status`, `limit`, `cursor` filters \| `await bb.ai.meetings.list({ status: 'done', limit: 50 })` \|
28
29	All methods return `{ data, error }` — always check `error` before using `data`.
30
31	---
32
33	## 3. Typical Flow
34
35	1. Configure webhook — call `manage_ai` with `action: "configure_meetings_webhook"`, passing `forward_url` pointing at a deployed Butterbase function or your own endpoint.
36	2. Start the bot — app calls `bb.ai.meetings.start(...)`. Status begins as `joining`.
37	3. Bot joins — status progresses `joining → waiting_room → in_call → recording`. The `bot.in_call_recording` event fires.
38	4. Call ends — bot status moves to `ended`, then `done` once artifacts are processed.
39	5. Webhook fires `bot.done` — your handler receives the event. The bot's `recordingUrl` and `transcriptUrl` are now populated.
40	6. Fetch artifacts — download the recording or transcript using the URLs from `bb.ai.meetings.get(id)`.
41	7. Write to substrate / database — store the transcript text, link to the relevant entity (deal, candidate, customer), or insert into a RAG collection.
42	8. Trigger downstream extract — run an LLM summarization, action-item extraction, or scoring job.
43
44	---
45
46	## 4. Webhook Handler Skeleton
47
48	Deploy this as a Butterbase function with `trigger: { type: "http", config: { method: "POST", auth: "none" } }`.
49
50	```typescript
51	export async function handler(req: Request, ctx: any): Promise<Response> {
52	const event = req.headers.get('x-bb-event');
53	const body = await req.text();
54	const payload = JSON.parse(body);
55
56	const botId: string = payload.data?.bot_id ?? payload.data?.id;
57
58	// Idempotency: skip if we already processed this bot + event combo.
59	const claimed = await ctx.idempotency.claim(`meeting:${botId}:${event}`);
60	if (!claimed) {
61	return new Response('duplicate', { status: 200 });
62	}
63
64	switch (event) {
65	case 'bot.done': {
66	// Artifacts are ready — store a reference or kick off extraction.
67	await ctx.db.query(
68	`UPDATE meeting_jobs SET status = 'done', bot_id = $1 WHERE external_ref = $2`,
69	[botId, payload.data?.metadata?.jobRef ?? botId]
70	);
71	break;
72	}
73	case 'transcript.done': {
74	// Transcript artifact URL is now available. Kick off downstream processing.
75	await ctx.db.query(
76	`INSERT INTO transcript_queue (bot_id, created_at) VALUES ($1, now())`,
77	[botId]
78	);
79	break;
80	}
81	case 'bot.fatal': {
82	await ctx.db.query(
83	`UPDATE meeting_jobs SET status = 'fatal' WHERE bot_id = $1`,
84	[botId]
85	);
86	break;
87	}
88	default:
89	// Other events (bot.in_call_recording, recording.done, transcript.failed) — handle as needed.
90	break;
91	}
92
93	return new Response('ok', { status: 200 });
94	}
95	```
96
97	Key points:
98	- Use `ctx.idempotency.claim(key)` — the meetings service may retry for up to 24 hours.
99	- Verify `x-bb-key-id` matches the first 16 characters of your current webhook secret to detect stale post-rotation events.
100	- Return `200` even for events you don't handle — a non-2xx response triggers a retry.
101
102	---
103
104	## 5. Cost-Aware Design
105
106	Before dispatching a bot in a user-pays context, call `estimateCost` and surface the projected charge:
107
108	```typescript
109	const { data: estimate } = await bb.ai.meetings.estimateCost({
110	duration

People who liked this also viewed

6 picks

Type

meetings

533stars40forks

View on GitHub

How to install meetings?

butterbase-ai/butterbase-skills/meetings

$npx -y skills add butterbase-ai/butterbase-skills --skill meetings

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Meeting Bots on Butterbase
2
3	Guide for building features that spawn meeting bots via `ctx.ai.meetings`. Covers the four SDK methods, webhook handling, idempotency, and cost-aware design.
4
5	---
6
7	## 1. When to Use
8
9	Reach for this skill when your feature involves any of:
10
11	- Meeting notetaker — bot joins a recurring standup or sales call and stores a transcript.
12	- Sales call enricher — join a customer call, wait for `bot.done`, then run an LLM over the transcript to populate a CRM record.
13	- Interview pipeline — record and transcribe candidate interviews, then score against a rubric.
14	- Voice-of-customer transcripts — capture support or discovery calls; feed chunks into a RAG collection for future search.
15
16	If you need only audio/video file upload without a live meeting join, use `manage_storage` instead.
17
18	---
19
20	## 2. The Four SDK Methods
21
22	\| Method \| What it does \| One-line example \|
23	\|--------\|-------------\|-----------------\|
24	\| `bb.ai.meetings.start(opts)` \| Spawns a bot and returns immediately with `status: "joining"` \| `const { data: bot } = await bb.ai.meetings.start({ meetingUrl, transcript: true })` \|
25	\| `bb.ai.meetings.get(id)` \| Fetches current status + artifact URLs \| `const { data } = await bb.ai.meetings.get(bot.id)` \|
26	\| `bb.ai.meetings.stop(id)` \| Removes the bot from the call early \| `await bb.ai.meetings.stop(bot.id)` \|
27	\| `bb.ai.meetings.list(opts)` \| Lists bots with optional `status`, `limit`, `cursor` filters \| `await bb.ai.meetings.list({ status: 'done', limit: 50 })` \|
28
29	All methods return `{ data, error }` — always check `error` before using `data`.
30
31	---
32
33	## 3. Typical Flow
34
35	1. Configure webhook — call `manage_ai` with `action: "configure_meetings_webhook"`, passing `forward_url` pointing at a deployed Butterbase function or your own endpoint.
36	2. Start the bot — app calls `bb.ai.meetings.start(...)`. Status begins as `joining`.
37	3. Bot joins — status progresses `joining → waiting_room → in_call → recording`. The `bot.in_call_recording` event fires.
38	4. Call ends — bot status moves to `ended`, then `done` once artifacts are processed.
39	5. Webhook fires `bot.done` — your handler receives the event. The bot's `recordingUrl` and `transcriptUrl` are now populated.
40	6. Fetch artifacts — download the recording or transcript using the URLs from `bb.ai.meetings.get(id)`.
41	7. Write to substrate / database — store the transcript text, link to the relevant entity (deal, candidate, customer), or insert into a RAG collection.
42	8. Trigger downstream extract — run an LLM summarization, action-item extraction, or scoring job.
43
44	---
45
46	## 4. Webhook Handler Skeleton
47
48	Deploy this as a Butterbase function with `trigger: { type: "http", config: { method: "POST", auth: "none" } }`.
49
50	```typescript
51	export async function handler(req: Request, ctx: any): Promise<Response> {
52	const event = req.headers.get('x-bb-event');
53	const body = await req.text();
54	const payload = JSON.parse(body);
55
56	const botId: string = payload.data?.bot_id ?? payload.data?.id;
57
58	// Idempotency: skip if we already processed this bot + event combo.
59	const claimed = await ctx.idempotency.claim(`meeting:${botId}:${event}`);
60	if (!claimed) {
61	return new Response('duplicate', { status: 200 });
62	}
63
64	switch (event) {
65	case 'bot.done': {
66	// Artifacts are ready — store a reference or kick off extraction.
67	await ctx.db.query(
68	`UPDATE meeting_jobs SET status = 'done', bot_id = $1 WHERE external_ref = $2`,
69	[botId, payload.data?.metadata?.jobRef ?? botId]
70	);
71	break;
72	}
73	case 'transcript.done': {
74	// Transcript artifact URL is now available. Kick off downstream processing.
75	await ctx.db.query(
76	`INSERT INTO transcript_queue (bot_id, created_at) VALUES ($1, now())`,
77	[botId]
78	);
79	break;
80	}
81	case 'bot.fatal': {
82	await ctx.db.query(
83	`UPDATE meeting_jobs SET status = 'fatal' WHERE bot_id = $1`,
84	[botId]
85	);
86	break;
87	}
88	default:
89	// Other events (bot.in_call_recording, recording.done, transcript.failed) — handle as needed.
90	break;
91	}
92
93	return new Response('ok', { status: 200 });
94	}
95	```
96
97	Key points:
98	- Use `ctx.idempotency.claim(key)` — the meetings service may retry for up to 24 hours.
99	- Verify `x-bb-key-id` matches the first 16 characters of your current webhook secret to detect stale post-rotation events.
100	- Return `200` even for events you don't handle — a non-2xx response triggers a retry.
101
102	---
103
104	## 5. Cost-Aware Design
105
106	Before dispatching a bot in a user-pays context, call `estimateCost` and surface the projected charge:
107
108	```typescript
109	const { data: estimate } = await bb.ai.meetings.estimateCost({
110	duration

People who liked this also viewed

6 picks

Type