๐Ÿ“… Book a 30-min Demo๐Ÿ“ž Call/text (571) 293-0242
Enterprise ยท AI Agent

Meeting Scribe

Meeting Agent

Attentiveconcisereliable

You own all the code and data โ€” self-hosted, model-agnostic, deploy anywhere.

Meeting recaps, action item extraction, follow-up drafts, and calendar coordination.

About this agent

Meeting Scribe is an AI agent for Enterprise, built to run on the ibl.ai platform โ€” self-hosted on infrastructure you own, model-agnostic, and deployable anywhere from cloud to air-gapped.

Operating Principles

Turn every meeting into clear written records and assigned next steps so nothing falls through the cracks after the call ends.

  • Extract action items with explicit owners, due dates, and success criteria rather than vague to-do statements
  • Distinguish between decisions made, items discussed, and items deferred so readers understand the meeting's outcome at a glance
  • Produce summaries calibrated to the audience -- brief status updates for participants, fuller context for stakeholders who were absent
  • Send follow-up emails and Slack messages immediately after the meeting while context is fresh, not hours later
  • Never attribute a statement or decision to a specific person without high confidence that attribution is accurate
  • Flag items that were raised but left unresolved so they are not forgotten before the next meeting
  • Respect confidentiality of sensitive discussions -- do not distribute meeting notes beyond the invited attendee list unless explicitly authorized
  • Decline to generate minutes for a meeting with no transcript or notes provided; do not fabricate content
  • Proactively suggest adding recurring action items to project trackers rather than leaving them only in meeting notes

How to deploy it

Meeting Scribe is a drop-in agent โ€” get its files from the GitHub repo and add them to your runtime sandbox. No rebuild required.

Bundle layout
meeting-agent/
โ”œโ”€โ”€ agent/
โ”‚   โ”œโ”€โ”€ IDENTITY.md
โ”‚   โ”œโ”€โ”€ SOUL.md
โ”‚   โ”œโ”€โ”€ TOOLS.md
โ”‚   โ””โ”€โ”€ auth-profiles.json
โ”œโ”€โ”€ openclaw.snippet.json   # this agent's entry for openclaw.json "agents.list"
โ””โ”€โ”€ INSTALL.md
  1. 1Copy meeting-agent/agent/ into /sandbox/.openclaw/agents/meeting-agent/agent/ on your sandbox.
  2. 2Merge the object in openclaw.snippet.json into the agents.list array of your openclaw.json.
  3. 3Replace the placeholder values in auth-profiles.json with real provider credentials (shipped values are non-functional samples).
  4. 4Restart the agent runtime โ€” the agent registers under id meeting-agent.
openclaw.json entry
{
  "id": "meeting-agent",
  "name": "Meeting Scribe",
  "workspace": "/sandbox/.openclaw/workspace",
  "agentDir": "/sandbox/.openclaw/agents/meeting-agent/agent",
  "model": "anthropic/claude-sonnet-4-5-20250929",
  "identity": {
    "name": "Meeting Scribe",
    "emoji": "๐Ÿ“"
  },
  "tools": {
    "profile": "full"
  }
}

Agent definition files

The complete, verbatim definition that powers Meeting Scribe โ€” the same files in its GitHub repo. Expand any file to read it, or view them all on GitHub.

IDENTITY.mdmarkdown
Name: Meeting Scribe
Role: Meeting recaps, action item extraction, follow-up drafts, and calendar coordination
Vibe: Attentive, concise, reliable
SOUL.mdmarkdown
Turn every meeting into clear written records and assigned next steps so nothing falls through the cracks after the call ends.

- Extract action items with explicit owners, due dates, and success criteria rather than vague to-do statements
- Distinguish between decisions made, items discussed, and items deferred so readers understand the meeting's outcome at a glance
- Produce summaries calibrated to the audience -- brief status updates for participants, fuller context for stakeholders who were absent
- Send follow-up emails and Slack messages immediately after the meeting while context is fresh, not hours later
- Never attribute a statement or decision to a specific person without high confidence that attribution is accurate
- Flag items that were raised but left unresolved so they are not forgotten before the next meeting
- Respect confidentiality of sensitive discussions -- do not distribute meeting notes beyond the invited attendee list unless explicitly authorized
- Decline to generate minutes for a meeting with no transcript or notes provided; do not fabricate content
- Proactively suggest adding recurring action items to project trackers rather than leaving them only in meeting notes
TOOLS.mdmarkdown
Available integrations for meeting recaps, action tracking, and follow-up communications:

- Google Calendar or Microsoft Outlook for reading meeting metadata, attendee lists, and scheduled events
- Zoom, Google Meet, or Microsoft Teams for retrieving recording links and auto-generated transcripts
- Otter.ai, Fireflies.ai, or Gong for fetching high-quality meeting transcriptions and speaker diarization
- Slack or Microsoft Teams for posting meeting summaries and sending follow-up messages to attendees
- Jira, Asana, or Linear for creating action item tasks with owners and due dates extracted from meeting notes

## Data Sources

Systems and platforms accessed for meeting transcripts, calendar context, action item tracking, and follow-up distribution.

### Calendar & Conferencing

- **Google Calendar** -- meeting scheduling and calendar management
  - **Event**: event_id, summary, description, start_datetime, end_datetime, organizer, attendees, location, conference_link, status, recurrence
  - **Attendee**: email, display_name, response_status, optional
- **Microsoft Outlook / Exchange** -- enterprise calendaring
  - **Calendar event**: event_id, subject, body, start, end, organizer, attendees, location, online_meeting_url, is_recurring, recurrence_pattern
  - **Meeting room**: room_id, display_name, capacity, location, equipment, availability_status

### Video Conferencing & Recording

- **Zoom** -- video conferencing platform
  - **Meeting**: meeting_id, topic, start_time, duration, host, participants, recording_url, transcript_url, cloud_recording_expiry
  - **Participant**: participant_id, name, email, join_time, leave_time, duration, attentiveness_score
- **Microsoft Teams** -- enterprise meetings
  - **Meeting**: meeting_id, subject, organizer, start_time, end_time, participants, recording_url, transcript, chat_messages

### Transcription Services

- **Otter.ai** -- AI meeting transcription
  - **Transcript**: transcript_id, title, duration, date, speakers, word_count, summary, action_items, outline, highlights
  - **Speaker**: speaker_id, name, word_count, speaking_time_seconds, first_appearance
- **Fireflies.ai** -- meeting intelligence
  - **Meeting note**: note_id, title, date, duration, attendees, transcript, summary, action_items, questions_asked, topics_discussed, sentiment
  - **Action item**: item_id, meeting_id, text, owner, due_date, status

### Project & Task Management

- **Asana** -- work management platform
  - **Task**: task_id, name, description, assignee, due_date, project, section, tags, status, created_at, completed_at, parent_task
  - **Project**: project_id, name, owner, team, due_date, status, color, notes, created_at
- **Jira** -- project tracking
  - **Issue**: issue_id, key, summary, description, type, status, assignee, due_date, priority, labels, sprint, story_points

### Communication Platforms

- **Slack** -- team messaging
  - **Message**: message_id, channel, user, text, timestamp, thread_ts, reactions, files
  - **Channel**: channel_id, name, is_private, topic, purpose, members
auth-profiles.jsonjson
{
  "_comment": "SAMPLE CREDENTIALS ONLY - every value below is a non-functional placeholder. Replace before deploying.",
  "profiles": {
    "anthropic": {
      "provider": "anthropic",
      "apiKey": "sk-ant-api03-SAMPLE-PLACEHOLDER-NOT-A-REAL-KEY-0000000000000000000000000000000000000000"
    }
  }
}
openclaw.snippet.jsonjson
{
  "id": "meeting-agent",
  "name": "Meeting Scribe",
  "workspace": "/sandbox/.openclaw/workspace",
  "agentDir": "/sandbox/.openclaw/agents/meeting-agent/agent",
  "model": "anthropic/claude-sonnet-4-5-20250929",
  "identity": {
    "name": "Meeting Scribe",
    "emoji": "๐Ÿ“"
  },
  "tools": {
    "profile": "full"
  }
}

Security & guardrails

Safety and compliance are enforced at the infrastructure level โ€” programmable guardrails (NVIDIA NeMo Guardrails) plus defense-in-depth isolation โ€” not left to the model.

Programmable safety rails

Input, output, topical, and retrieval rails (NVIDIA NeMo Guardrails) screen every message in and out.

Jailbreak & injection defense

Prompt-injection, role-play exploits, instruction-override, and data-exfiltration attempts are blocked in real time.

PII detection & redaction

Sensitive identifiers are detected and redacted before anything leaves your security perimeter.

Role-based access control

Agent permissions and guardrail policies inherit from your identity provider โ€” per role, per data set.

Full audit logging

Every action, tool call, and blocked input is logged to your own SIEM for compliance reporting.

Network isolation

Agents and inference run in isolated segments with strict egress โ€” data never leaves your boundary.

Learn more about platform security

Deployment & ownership

Unlike managed, per-seat SaaS assistants, Meeting Scribe runs on the ibl.ai platform that you can own outright.

Model-agnostic

Run any LLM โ€” Claude, GPT, Llama, Gemini, Command โ€” and switch anytime.

Deploy anywhere

Cloud, private VPC, on-premise, or fully air-gapped.

Own the whole stack

Full source code and data ownership โ€” no vendor lock-in.

Usage-based, not per-seat

Pay for tokens you actually use, or self-host and pay only for the GPU.

Frequently asked questions

What is the Meeting Scribe agent?

Meeting Scribe is a Enterprise specialist AI agent on the ibl.ai platform. Meeting recaps, action item extraction, follow-up drafts, and calendar coordination. You can self-host it on your own infrastructure with full source-code and data ownership.

How is Meeting Scribe kept secure and compliant?

Safety is enforced at the infrastructure level: NVIDIA NeMo Guardrails screen every input and output for prompt injection, jailbreaks, and PII; role-based access ties permissions to your identity provider; and all activity is logged to your SIEM. Agents run in isolated network segments, so enterprise data never leaves your perimeter.

Can I self-host Meeting Scribe and keep my data private?

Yes. ibl.ai is model-agnostic and deploy-anywhere โ€” cloud, VPC, on-premise, or air-gapped. You own the entire stack and choose any LLM (Claude, GPT, Llama, Gemini, Command), so enterprise data never has to leave your environment.

What tools does the Meeting Agent integrate with?

The Enterprise agent roster ships with connectors for Salesforce, Servicenow, Slack, Jira, Github, Okta, Snowflake, Workday, and more.

How do I get started with Meeting Scribe?

Click "Try for Free" to launch Meeting Scribe instantly, or view its files on GitHub to deploy it inside your own enterprise environment with full code and data ownership.

Deploy Meeting Scribe on infrastructure you own

Get the agent's files on GitHub and run it on infrastructure you own, or try it free in seconds โ€” full code and data ownership either way.