Stand-up Meetings

AI briefings for human decision-makers.

Overview

A stand-up meeting is a 1:1 conversation between you and an AI agent about an initiative. The AI presents its findings, reports on progress, surfaces open questions, and asks you for decisions and clarifications.

This is NOT a team stand-up. It's an AI briefing you, the decision-maker.

URL: https://argus.meetrhea.com/initiatives/:id/standup

How It Works

1. AI Presents

The agent briefs you on current state:

• "Here's what I found..."
• "Current state of the initiative..."
• "Research uncovered..."

2. AI Asks

The agent surfaces decisions it cannot make alone:

• "Should we pivot on X?"
• "I need a decision on Y"
• "What's your priority for Z?"

3. You Decide

You give direction, clarify intent, answer questions.

4. AI Records

The agent captures your decisions:

• Creates artifacts attached to the initiative
• Updates criteria confidence levels
• Logs decisions with rationale

Research On-Demand

At any point, trigger research when questions arise:

• You ask something the AI can't answer from context
• AI identifies a knowledge gap while presenting
• Either of you says "let me look that up"

Like having a librarian on-call who can pull relevant information mid-conversation.

What the AI Presents

• Initiative state - Progress since last stand-up, criteria confidence
• Findings - What research discovered, patterns identified
• Open questions - Decisions only you can make

What the AI Asks

Category	Example Questions
Direction	"Two approaches emerged. Which aligns with your goals?"
Priority	"Three tickets are ready. Which should I tackle first?"
Clarification	"You mentioned X last time. Did you mean A or B?"
Approval	"I drafted this decision record. Does it capture your intent?"

Agent Role

The AI is:

• Presenter - Reports findings, progress, blockers
• Question-asker - Surfaces decisions needed from you
• Researcher - Looks things up when questions arise
• Recorder - Captures decisions and updates initiative state

The AI is NOT:

• A facilitator of team meetings
• Asking "what did you do yesterday?"
• Running a round-robin format

Artifacts Generated

Type	When Created
meeting_notes	Always - summary of what was discussed
decision_record	When you made a decision
updated criteria	When confidence levels change

You review and approve before saving.

Integration

Claudebox Agent SDK

Stand-ups use the Claude Agent SDK via Claudebox for:

• Streaming conversation responses
• Full MCP tool access (Janus for initiative context)
• Session persistence

Research Tools

• janus_second_opinion - Quick external validation
• janus_brain_thread_* - Ongoing research threads

Artifact Creation

• janus_artifact_create - Meeting notes, decision records
• janus_criteria_update - Confidence level changes
• janus_decision_log - Record decisions with rationale

API

# Start stand-up session
POST /api/initiatives/:id/standup

# Response: SSE stream
data: {"type": "message", "role": "assistant", "content": "..."}
data: {"type": "research", "query": "...", "results": [...]}

# Send your response
POST /api/initiatives/:id/standup/message
{
  "content": "Your response here"
}

# Trigger research
POST /api/initiatives/:id/standup/research
{
  "query": "What approaches do other teams use for..."
}

# End and generate artifacts
POST /api/initiatives/:id/standup/end
{
  "generate_artifacts": true
}

When to Use Stand-ups

• Initiative has new research to present
• Decisions are needed to move forward
• Criteria confidence needs validation
• You want a status briefing on progress

Comparison

This IS	This is NOT
AI briefing human	Human team stand-up
AI asking for decisions	AI facilitating discussion
1:1 human-AI conversation	Round-robin format
AI as reporter/researcher	AI as scrum master
On-demand, when needed	Daily ritual