AI Team Collab — Command Center v6

29 models loaded · 8:48 AM

Select the models for the roles you want to use.

How to Configure Roles

Each role card has controls to configure your AI team:

● Warm Indicator: — Click to pre-load the model. Green = ready, yellow = loading, gray = cold.
☐ Add to Workspace: — Include this role in your current mission's active team.
📋 Brief: — View or edit the role's system prompt and instructions.
🪑 Seat at Table: — Add this role to the active conversation. Seated roles respond to messages.
💬 Chat History: — View past conversations for this role. Coming soon.
🧠 Assigned Memory: — Link memory namespaces for persistent knowledge. Coming soon.

Tip: Use the model dropdown on each card to assign a specific AI model to that role.

Role × Abilities Matrix

Role	🔧	👁️	📊	🧠	🖼️	🎬	📏
Commander	✅	➖	➖	💛	➖	➖	💛
Coder	✅	➖	➖	⚙️	➖	➖	➖
Reviewer	💛	➖	➖	✅	➖	➖	💛
Architect	💛	➖	➖	✅	➖	➖	💛
Researcher	✅	💛	✅	✅	➖	➖	✅
Security	✅	➖	➖	✅	➖	➖	💛
Tester	✅	➖	➖	💛	➖	➖	➖
Docs	💛	➖	➖	➖	➖	➖	✅
Reporter	💛	➖	➖	➖	➖	➖	✅
ImageGen	💛	💛	➖	➖	✅	➖	➖
VideoGen	💛	💛	➖	➖	➖	✅	➖
UX Designer	💛	✅	➖	💛	➖	➖	💛
DevOps	✅	➖	➖	💛	➖	➖	➖
Analyst	✅	💛	✅	✅	➖	➖	💛

✅ Required 💛 Preferred ⚙️ Per-preset ➖ Not needed

🔧 Tools · 👁️ Vision · 📊 Embed · 🧠 Thinking · 🖼️ GenImg · 🎬 GenVid · 📏 LongCtx

Work Mode:

🚀 Quick Start ~30 seconds

☐ Login

☐ Choose TEAM PRESET

☐ Click Apply

☐ Send first message

☐ Check AI Activity

Commander: ⏳ checking...

0/5 complete

AIteamCollab Command Center

Loading mission data...

Phase: — • Progress: —

Work Role: —

📜 Workflow Recipe

📦 Team Preset

🔄 Workflow Mode

Tandem Mode

Mode is locked per mission. Changes apply to the next mission only.

📋 Role Briefs

Model: —

No brief defined yet.

👥 Active Crew (0 seated)

Active Commander

—

No roles seated

Plan OK

Code Review

Security OK

📋 Tasks / ToDo

Loading tasks...

RAM & Models Updated: --

RAM:

--/-- GB --

Models: -- warm

Queue: --

Services:

API WS Ollama

▼ Click to expand model list

📦 Git 💬 Gitea 📚 Docs

Mission: —

Agent thinking traces will appear here...

Preset: None • Ask Team: OFF

Welcome to AITeamCollab.com's Commander Codey!

I can help you Write Stories, Reports, Website Content, Code, and Build Apps with a Multi-Model Agent Workflow using free models, or the popular cloud models also.

First, tell me:
What level would you like to work as with Commander Codey?

Don't show this message again.

You can change this option at any time in the Settings Panel

[Skip for now, just let me type]

System Logs Auto

Loading logs...

System Notices

No notices yet...

Models

Live

Models

Models · Loading... Show Inactive

Loading models...

Loading cloud models...

Idle · Ready for input

Waiting for events...

Verbose Pin

Ctrl+. to toggle Enter to execute

Settings

admin

Workspace home

Providers & Models

Providers

Default Documentation Model Model used for artifact generation (Handoff Docs). Saved to server config.

RAM Safety Mode

Strict (block risky combinations) Allow with Warning (permit but show alerts)

Controls whether over-budget model combinations are blocked or allowed with warnings.

Team & Pipeline

Team Workflow Mode

Parallel (Commander + Team respond simultaneously) Sequential (Commander → Coder → Reviewer chain)

When Ask Team is ON: Parallel fans out all roles at once; Sequential chains them in order.

Warmup Settings

Follow team seats for warmup

When enabled, models assigned to team seats are auto-warmed.

Manual warmup pins Comma-separated Ollama model IDs to always keep warm (cloud models rejected).

Model Filtering

Show all models in role dropdowns

Limit max size GB

Filter role dropdowns to show only models that fit your machine.

Mission Defaults

Default Deployment Target

Default App Blueprint

Work Commands

Default Target

Show recent commands Per-workspace history Allow custom commands

Storage & Export

Deployment History ?

Keep last N deployments

Archive format Format used when archiving deployment history.

Enable Google Drive backup

Google Drive folder ID Enter the folder ID from the Drive URL, not the full URL.

Google Drive: checking...

Mission Governance

Guidance Level Controls onboarding and hint verbosity.

Auto-update docs after missions

Budget Enforcement

Pause on budget threshold

Checkpoint Settings

Checkpoints enabled

Checkpoint frequency (seconds)

10s30s300s

Save on step complete

Save on state change

Mission Budget Defaults Default limits for each mission type. Click to expand/collapse.

PLAN IT

Max duration (min) [1-600]

Max tool calls [1-10000]

Max file edits [1-2000]

Max tokens [1K-5M]

BUILD IT

Max duration (min) [1-600]

Max tool calls [1-10000]

Max file edits [1-2000]

Max tokens [1K-5M]

TEST IT

Max duration (min) [1-600]

Max tool calls [1-10000]

Max file edits [1-2000]

Max tokens [1K-5M]

SHIP IT

Max duration (min) [1-600]

Max tool calls [1-10000]

Max file edits [1-2000]

Max tokens [1K-5M]

Current Applied Budgets

Read-only view of budgets embedded in mission definitions (missions.json). These are the actual limits applied at mission start.

Mission
Duration
Tools
Tokens
Loading...

Security & Permissions

Tools / Gates

Per-role file capabilities (R=read, W=write, X=execute)

Role	R	W	X	Mode

Heavy Build Protection

When enabled, applies extra safety checks before large-scale operations (npm install, pip install, builds, etc.). Enable heavy build protection

Execute Allow-List

Commands AI can execute during missions. Toggle to enable/disable. Per-role overrides inherit global unless explicitly set.

Global Commands These apply to all roles unless overridden.

Per-Role Overrides Roles inherit global commands. Add overrides to enable extra commands or disable inherited ones for a specific role.

Select role to view/edit overrides

No overrides for this role

Warmup Settings

Warmup Mode Controls proactive model warming. Sequential2/3 are placeholders for future multi-target behavior.

Models to auto-warm (one per line) Only these models will be auto-warmed by the daemon. Empty = allow all (minus blacklist).

Warmup timeout (seconds) This timeout is the daemon per-model warm request timeout, not keepalive. Higher values may be needed for larger vision models on CPU.

Per-model warmup timeouts You can set a custom warmup timeout (in seconds) for any model. Leave a model blank to use the global timeout. Use the filter to quickly find models (filtering only changes what you see, not what gets saved). Override the global timeout for specific models. Leave empty to use global timeout.

Warmup Snapshot —

Degradation Defaults

Enabled Require Preflight Allow Asymmetric Fallback Log Artifacts

Default Mode

Floor Class

Max Tiers (1-10)

Max Events (1-100)

Latest Tandem Degradation —

Observability

Live Actions Verbosity

Default Verbose

Default shows high-value pipeline events. Verbose shows everything.

Thinking Tab Content

Reasoning Metadata Only

Reasoning includes content summaries. Metadata shows timestamps and kinds only.

Show Dialogue (Chat) Tab

When off, the Chat subtab reverts to role status view.

Welcome to AITeamCollab.com's Commander Codey!

System Notices

⚠ Force Quit (Last Resort)

Login required

Welcome to AITeamCollab.com's Commander Codey!

System Notices

Artifact Preview

File Change Approval

Mission Recovery