Vexa MCP can request bots to join meetings, stop bots, parse meeting links, read transcripts, manage recordings, create transcript share links, send and read meeting chat, make the bot speak, share screen content, update bot config, connect calendars, and configure webhooks., Meeting URLs, IDs, passcodes, bot names, platform names, participant data, recordings, transcript share links, and API keys require careful handling., Public transcript share links and recording download URLs should be short-lived and shared only with approved recipients., Interactive bot tools can visibly act in live meetings; require human approval before using chat, speech, or screen-sharing tools., The service README includes a DoD table that marks several MCP checks as untested, so validate the exact deployment path before production use.
Privacy notes
Audio, transcripts, speaker labels, meeting metadata, chat messages, recordings, screenshots, participant names, calendar events, webhook payloads, meeting notes, API keys, and tool results may be visible to the MCP client and model provider., Hosted Vexa sends meeting data to Vexa-operated infrastructure; self-hosted deployments still may use hosted transcription tokens unless fully self-hosting transcription., Browser bot sessions can persist authenticated browser state through storage such as S3-compatible backends., Logs, databases, object storage, share links, webhook receivers, and AI prompts can retain sensitive meeting content after a call ends.
Author
Vexa AI
Submitted by
oktofeesh1
Claim status
unclaimed
Last verified
2026-06-05
Decision playbook
Review trust signals before you adopt
Signals are present but mixed. Use the checklist below to confirm the source and operational safety for your environment.
Compare context
Selected
0
Current score
63
Baseline
—
Delta
No baseline selected
No major trust-signal divergence detected in the current selection.
Source and provenance checks
Needs review
Confirm ownership and provenance before trusting install instructions.
Source link availableRequired
Open the canonical repository and verify ownership.
Done
Source provenance statusRequired
Marked as source-backed.
Done
Metadata reviewed
No reviewed flag detected in metadata.
Pending
Safety and privacy checks
Complete
Validate risk disclosures before installation or API wiring.
Safety notes presentRequired
Review the listed safety guidance before running commands.
Done
Privacy notes presentRequired
Review data handling notes before connecting accounts or secrets.
Done
Trust level risk gateRequired
Trust level does not block evaluation.
Done
Package and install checks
Needs review
Check package metadata and artifact integrity signals.
Install payload available
Install or copy payload is available for review.
Done
Package verification flag
No package verification flag provided.
Pending
Checksum metadata
No checksum provided for downloaded artifact.
Pending
Compare-driven decision checks
Needs review
Use compare context to validate trade-offs before adoption.
Compare tray has multiple entries
Add at least one more entry to compare trust differences.
5 safety and 4 privacy notes across 4 risk areas. Review closely: credentials & tokens, network access.
4 areas
SafetyNetwork accessVexa MCP can request bots to join meetings, stop bots, parse meeting links, read transcripts, manage recordings, create transcript share links, send and read meeting chat, make the bot speak, share screen content, update bot config, connect calendars, and configure webhooks.
SafetyCredentials & tokensMeeting URLs, IDs, passcodes, bot names, platform names, participant data, recordings, transcript share links, and API keys require careful handling.
SafetyNetwork accessPublic transcript share links and recording download URLs should be short-lived and shared only with approved recipients.
SafetyGeneralInteractive bot tools can visibly act in live meetings; require human approval before using chat, speech, or screen-sharing tools.
SafetyLocal filesThe service README includes a DoD table that marks several MCP checks as untested, so validate the exact deployment path before production use.
PrivacyCredentials & tokensAudio, transcripts, speaker labels, meeting metadata, chat messages, recordings, screenshots, participant names, calendar events, webhook payloads, meeting notes, API keys, and tool results may be visible to the MCP client and model provider.
PrivacyCredentials & tokensHosted Vexa sends meeting data to Vexa-operated infrastructure; self-hosted deployments still may use hosted transcription tokens unless fully self-hosting transcription.
PrivacyCredentials & tokensBrowser bot sessions can persist authenticated browser state through storage such as S3-compatible backends.
PrivacyNetwork accessLogs, databases, object storage, share links, webhook receivers, and AI prompts can retain sensitive meeting content after a call ends.
Safety notes
Vexa MCP can request bots to join meetings, stop bots, parse meeting links, read transcripts, manage recordings, create transcript share links, send and read meeting chat, make the bot speak, share screen content, update bot config, connect calendars, and configure webhooks.
Meeting URLs, IDs, passcodes, bot names, platform names, participant data, recordings, transcript share links, and API keys require careful handling.
Public transcript share links and recording download URLs should be short-lived and shared only with approved recipients.
Interactive bot tools can visibly act in live meetings; require human approval before using chat, speech, or screen-sharing tools.
The service README includes a DoD table that marks several MCP checks as untested, so validate the exact deployment path before production use.
Privacy notes
Audio, transcripts, speaker labels, meeting metadata, chat messages, recordings, screenshots, participant names, calendar events, webhook payloads, meeting notes, API keys, and tool results may be visible to the MCP client and model provider.
Hosted Vexa sends meeting data to Vexa-operated infrastructure; self-hosted deployments still may use hosted transcription tokens unless fully self-hosting transcription.
Browser bot sessions can persist authenticated browser state through storage such as S3-compatible backends.
Logs, databases, object storage, share links, webhook receivers, and AI prompts can retain sensitive meeting content after a call ends.
Prerequisites
Node.js and npx available for the `mcp-remote` client bridge.
Vexa API key for hosted Vexa or an approved self-hosted Vexa deployment.
Meeting organizer permission and participant consent before joining, recording, transcribing, chatting, speaking, or screen-sharing through a bot.
Platform policy review for Google Meet, Microsoft Teams, and Zoom bot participation.
Storage, transcript retention, recording retention, webhook, calendar, and share-link policies reviewed before use.
Vexa MCP Server exposes Vexa meeting bot and transcription capabilities through
MCP. It lets Claude request bots for Google Meet, Microsoft Teams, and Zoom,
fetch transcripts, manage recordings, create transcript share links, control
meeting chat or bot speech, configure webhooks, and work with meeting metadata.
The MCP service is implemented as a FastAPI proxy over Vexa's API gateway. The
repo supports hosted Vexa API usage, Vexa Lite for a single-container
self-hosted deployment, Docker Compose for multi-service deployments, and
service-specific MCP docs for client setup.
These sources were reviewed on 2026-06-05. Prefer the live repository,
README, MCP docs, MCP service source, service config, requirements, deployment
guides, and Vexa docs for current tool names, deployment paths, API key handling,
platform support, transcript APIs, WebSocket behavior, and service maturity.
Features
Parse meeting links into platform, meeting ID, and passcode fields.
Request or stop meeting bots for Google Meet, Microsoft Teams, and Zoom.
Fetch live or post-meeting transcripts with speaker attribution.
Create temporary transcript share links and meeting bundles.
List, fetch, download, and delete recordings.
Send or read meeting chat messages through an interactive bot.
Make the bot speak with text-to-speech or stop speaking.
Configure recording defaults, webhooks, calendar connection, and meeting metadata.
Use a hosted Vexa endpoint or self-host Vexa Lite, Docker Compose, or Helm deployments.
Installation
For the quickest self-hosted evaluation, the repository documents Vexa Lite:
git clone https://github.com/Vexa-ai/vexa.git
cd vexa
make lite
For MCP clients, configure mcp-remote with the hosted or self-hosted MCP
endpoint and an API key:
Use the hosted Vexa endpoint only when your meeting data handling policy allows
it. For self-hosted deployments, use the MCP service endpoint from your own Vexa
environment.
Use Cases
Ask Claude to join a meeting as an approved notetaking bot.
Fetch the current transcript during a live call.
Build a post-meeting summary from transcripts, recordings, and meeting metadata.
Send a short approved chat message from the bot.
Trigger post-meeting transcription of a recording.
Configure webhooks for meeting events.
Connect calendar data for upcoming meeting workflows.
Use self-hosting when meeting data must remain under your infrastructure.
Safety and Privacy
Meeting automation is high-trust. Obtain organizer permission and participant
consent before joining, recording, transcribing, chatting, speaking, sharing
screen content, or creating share links. Review Google Meet, Microsoft Teams,
Zoom, and local legal requirements before use.
Treat transcripts, recordings, share links, browser sessions, chat messages,
calendar events, and webhook payloads as sensitive meeting data. Hosted Vexa may
send data to Vexa-operated infrastructure, while self-hosted deployments still
need careful storage, retention, token, logging, and transcription-service
configuration.
Duplicate Check
No Vexa-ai/vexa entry, Vexa MCP Server entry, or matching source URL was found
in content/mcp.
Show that Vexa MCP Server is listed on HeyClaude. Paste this Markdown into your README — it renders the badge and links back to this page.
[](https://heyclau.de/entry/mcp/vexa-mcp-server)
How it compares
Vexa MCP Server side by side with 3 alternatives on trust, install, platform support, and disclosed safety notes — all from reviewed registry metadata.
3 trust signals differ across this comparison (Package trust, Source provenance, Submitter).
MCP server for Microsoft Teams channel workflows, including reading threads, posting new threads, replying to threads, mentioning members, and listing team members.
✓Vexa MCP can request bots to join meetings, stop bots, parse meeting links, read transcripts, manage recordings, create transcript share links, send and read meeting chat, make the bot speak, share screen content, update bot config, connect calendars, and configure webhooks.
Meeting URLs, IDs, passcodes, bot names, platform names, participant data, recordings, transcript share links, and API keys require careful handling.
Public transcript share links and recording download URLs should be short-lived and shared only with approved recipients.
Interactive bot tools can visibly act in live meetings; require human approval before using chat, speech, or screen-sharing tools.
The service README includes a DoD table that marks several MCP checks as untested, so validate the exact deployment path before production use.
✓Connect only meetings and workspaces intended for AI review, especially when transcripts include customers or internal discussions.
✓MCP Teams Server can start new channel threads, reply to existing threads, mention members, read thread replies, list channel threads, look up members by name, and list team members.
Posting or replying in Microsoft Teams can notify real teammates and create persistent workplace records.
Mentioning a member can trigger direct notifications, so require human review before allowing model-generated mentions.
Channel read tools can expose existing conversations, replies, message IDs, pagination cursors, and team member context to the model.
The required Microsoft Graph permission includes ChannelMessage.Read.All resource-specific consent, and Teams app permissions include message send/read and member read scopes.
Use a dedicated bot application, scoped Team ID, scoped Channel ID, least-privilege permissions, and explicit approval for every write operation.
Launch with `uvx --from git+https://github.com/InditexTech/mcp-teams-server mcp-teams-server` or from a verified checkout; do not use plain `uv run mcp-teams-server` from untrusted workspaces because local projects, virtual environments, or PATH entries can hijack the executable while receiving Teams secrets.
✓The unified server can discover and execute many ACI.dev managed functions, so restrict permissions before connecting it to an autonomous agent.
App-specific servers should be preferred when a workflow only needs a small set of integrations.
Connected apps may read, create, update, delete, or send data depending on the selected functions and linked account permissions.
Require human approval before email, calendar, CRM, ticketing, infrastructure, billing, deployment, or account-administration actions.
Privacy notes
✓Audio, transcripts, speaker labels, meeting metadata, chat messages, recordings, screenshots, participant names, calendar events, webhook payloads, meeting notes, API keys, and tool results may be visible to the MCP client and model provider.
Hosted Vexa sends meeting data to Vexa-operated infrastructure; self-hosted deployments still may use hosted transcription tokens unless fully self-hosting transcription.
Browser bot sessions can persist authenticated browser state through storage such as S3-compatible backends.
Logs, databases, object storage, share links, webhook receivers, and AI prompts can retain sensitive meeting content after a call ends.
✓Meeting transcripts, summaries, speaker names, attendee metadata, and conversation details may be sent through model context.
✓Teams channels can contain customer data, employee names, emails, incident details, roadmap discussions, support conversations, credentials, links, and confidential attachments or references.
Tool calls and logs can expose TEAMS_APP_ID, tenant IDs, team IDs, channel IDs, thread IDs, message IDs, member names, member emails, message bodies, replies, mentions, and pagination cursors.
TEAMS_APP_PASSWORD, client secrets, tenant IDs, bot configuration, and `.env` files should stay out of prompts, issues, logs, screenshots, and committed files.
Claude-generated messages may become discoverable records under organization retention, compliance, and eDiscovery policies.
✓Tool names, function arguments, prompts, linked account identifiers, execution results, logs, and connected-service data may pass through ACI.dev, the selected third-party apps, the MCP client, and the model provider.
API keys and linked-account credentials must be protected in client config, environment files, logs, screenshots, and shared prompts.
Review ACI.dev account, tenant, retention, audit-log, and OAuth settings before routing production or customer data through the server.
Prerequisites
Node.js and npx available for the `mcp-remote` client bridge.
Vexa API key for hosted Vexa or an approved self-hosted Vexa deployment.
Meeting organizer permission and participant consent before joining, recording, transcribing, chatting, speaking, or screen-sharing through a bot.
Platform policy review for Google Meet, Microsoft Teams, and Zoom bot participation.
Fireflies.ai account (sign up at https://fireflies.ai if needed)
Fireflies API key (obtain from Fireflies settings/API section)
OAuth authentication support (for workspace access)
HTTP transport support (remote MCP server at https://api.fireflies.ai/mcp)
Python 3.10 and uv/uvx available to the MCP client runtime.
Microsoft Teams account and a Team/Channel where the bot app is installed.
Microsoft Entra ID application registration with client secret credentials.
Azure Bot registration connected to the Microsoft Teams channel.
ACI.dev account, API key, and linked account owner ID.
Connected app credentials or linked accounts for the tools the agent should use.
MCP client that can launch Python package commands with environment variables.
Clear allowlist of apps, functions, and user accounts before exposing write-capable tools.
Install
git clone https://github.com/Vexa-ai/vexa.git && cd vexa && make lite