Agent Configuration

Agents are AI employees on TeamDay — specialized team members you install, start with, or create for specific roles. This guide covers the settings available when configuring an agent.

Configuration depends on the agent type:

• Ready Agents: configure credentials, context, scope, schedule, and approvals. Teamday manages the global prompt, rubric, and upgrades.
• Teamday Agents: chat with them and grant local context. Teamday manages their core identity and behavior.
• Private Agents: fully editable agents owned by your organization.

---

Agent Fields

Private Agents have the following configurable fields. Ready Agents and Teamday Agents expose only the fields that are safe for their mutability level.

---

Equipment

Agents have four types of equipment, each managed via a dedicated tab in the agent detail panel:

Tools

Built-in Claude tools that every agent has access to by default:

Tools are universal — the same set is available whichever folder the agent is working in. Use disabledTools to restrict specific tools, or allowedTools to set an explicit allowlist.

Skills

Reusable capabilities attached to the agent via skillIds. Skills provide domain expertise, workflows, and automation scripts.

{
  "skillIds": ["core:research-assistant", "core:data-analyst"]
}

Skills belong to the agent and travel with it into every folder it works in. There are no per-folder skill sets.

See Skills for how to build and manage skills.

Browser Computer is a built-in hosted runner skill for web browsing with a persistent browser profile. Use it when an agent needs screenshots, JavaScript-heavy pages, or authenticated website workflows. See Browser Computer.

MCPs

External integrations attached to the agent via mcpInstanceIds. MCP (Model Context Protocol) servers connect agents to external services like Google Analytics, GitHub, Slack, etc.

{
  "mcpInstanceIds": ["mcp-instance-id-1", "mcp-instance-id-2"]
}

MCP connections belong to the agent and are available wherever it works. Connect an MCP server to the agent once; it does not need to be repeated per folder.

See MCP Servers for available integrations.

Subagents

Other agents this agent can delegate work to, attached via subagentIds. Subagents run isolated tasks and report results back.

{
  "subagentIds": ["agent-id-1", "agent-id-2"]
}

Subagents can also be defined as markdown files in a .claude/agents/ directory inside the working directory. The agent discovers both its configured subagents and any filesystem-based subagents found there.

Equipment Scope

An agent's equipment (tools, skills, MCP connections, subagents) is the same in every chat, job, and mission. The only thing that changes between runs is the working directory: which folder the agent reads and writes by default.

---

Models

Available Models

TeamDay supports models from multiple providers:

Claude (Anthropic) — Claude Code harness:

OpenAI — Codex harness:

Gemini (Google) — Google Antigravity harness:

Gemini models bill through the connected Google Antigravity account, not per token. Gemini 3 Flash and Gemini 3.5 Flash have left the selector; saved selections map to the same thinking level of Gemini 3.8 Flash.

Qwen Code (Alibaba Cloud):

Qwen Code models require a connected Qwen Code provider account (Alibaba Cloud Model Studio or an OpenAI-compatible endpoint). Configure it in Settings → Connected Providers → Qwen Code.

Setting the Model

CLI:

teamday agents create --name "Fast Bot" --model claude-haiku-4-5 --system-prompt "..."

API:

{
  "name": "Fast Bot",
  "model": "claude-haiku-4-5",
  "systemPrompt": "..."
}

---

Visibility

Visibility controls who can discover your agent. It does not control which folders the agent works in — that's chosen when someone starts a chat or mission, within the agent's working-folder limits.

Public agents get a marketing page at teamday.ai/team/{slug} with their description, FAQ, use cases, and a "Add to Team" button.

---

System Prompt

For Private Agents, the systemPrompt field is the core of your agent's behavior. Ready Agents and Teamday Agents have Teamday-managed core prompts; configure their local context and credentials instead of rewriting their identity.

Key tips:

• Start with identity: "You are [Name], a [role]"
• Define behavior: what the agent should do and how
• Set constraints: what the agent should not do
• Be specific: vague prompts produce vague results

---

State of Mind

Open an agent's Overview to read its current business assessment: how its area is doing, what is working, what needs attention, near-term priorities, and longer-term direction. The assessment shows when it was published, the agent's confidence, and a Review source link to the run behind it. Confidence is the agent's own judgment.

What changed recently appears when the agent reports a change in results, understanding, or priorities. It explains what changed and why it matters; older assessments without this information do not show an empty section.

Missions and assigned tasks ask agents to update this assessment using available evidence. A later procedural report does not erase the last assessment. If no assessment has been published, the profile says so. Priorities describe intended work; they are not proof that tasks have been assigned or completed.

When you chat with an agent, its latest published assessment, including any recent changes, is supplied as context for its reply. You can discuss what you see on the profile directly. The assessment retains its publication date and confidence; it is previous judgment to reconsider when new evidence arrives.

Working now appears when work is active. Expand Mission history and schedule for assigned work, scheduled runs, and operational reports.

Working Folders

Agents belong to your organization and are not tied to any folder. Each chat, job, and mission has its own working directory, chosen when it is created; the agent brings the same instructions, skills, and MCP connections to every one of them.

Three settings shape where an agent works:

How a Run Uses Them

Relative paths in instructions resolve from the working directory. Agents can read, search, and publish Knowledge documents in folders they may access, unless their default Teamday connection has been explicitly disabled. They can also choose a permitted folder for delegated work or a new mission through the available Teamday tools; doing so does not change the current run's working directory. Agent memory persists as work moves between folders.

When to Restrict Paths

Restrict read or write paths when:

• Your organization has multiple teams and one team's agent should not read another team's folders.
• You are running an agent with elevated capabilities (e.g., an admin agent) and want to minimize its blast radius.
• You are setting up a customer-facing agent that should only see one customer's folder.

Leave both empty (the default) when the agent is designed to synthesize information across your entire organization.

Configuring Working Folders

CLI:

teamday agents create \
  --name="Support Analyst" \
  --role="Support analyst" \
  --system-prompt="..." \
  --default-cwd=/support \
  --read-paths=/support,/knowledge \
  --write-paths=/support

Web app: open Agent Detail for any Private Agent, click Edit agent (or Configure), expand Advanced, and set the default working directory and read/write paths. Save; the new limits apply to all subsequent runs.

---

Platform MCP Tools

TeamDay provides built-in MCP tools that all agents can use:

See Platform Tools for detailed documentation.

---

Updating Configuration

Private Agent fields can be updated after creation. Ready Agents and Teamday Agents reject protected prompt, identity, rubric, and tool-policy edits.

CLI:

teamday agents update <private-agent-id> --model claude-opus-4-6 --system-prompt "New instructions..."
teamday agents update <id> --skills "core:research-assistant,new-skill"

API:

curl -X PATCH "https://cc.teamday.ai/api/v1/agents/<id>" \
  -H "Authorization: Bearer $TEAMDAY_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"model": "claude-opus-4-6"}'

Only provided fields are updated. Omitted fields remain unchanged. Protected Teamday-managed fields return an authorization error instead of mutating the agent product.

---

Next Steps

• Prompts & Instructions — Write better system prompts
• Tools — Deep dive into available tools
• Browser Computer — Give agents a persistent hosted browser
• Skills — Build and install reusable capabilities
• Folders and Files — Where agents read and write