Skip to content

Obsy Agent Teams


Obsy Agent Teams is a team-oriented Agent workspace. You can create Agents with different responsibilities, configure Skills, MCP services, and external messaging platforms for them, and then use tasks and scheduled tasks to enable Agents to continuously participate in daily team work.

Obsy Agent Teams is suitable for the following scenarios:

  • Create dedicated Agents for responsibilities such as R&D, operations, testing, data analysis, observability troubleshooting, and knowledge organization.
  • Control the visibility and availability of Agents by member or role.
  • Solidify Agents' working methods through Skills, such as code review, test generation, document organization, browser automation, etc.
  • Connect to external tools or internal systems through MCP services, and assign available capabilities according to Agent responsibilities.
  • Integrate Agents into external IMs like WeChat, DingTalk, and Lark, enabling team collaboration within familiar messaging platforms.
  • Create tasks or scheduled tasks, allowing Agents to continuously execute work around clear objectives, record the process, and accumulate results.

It is recommended to first create Agents with clear responsibility boundaries, then configure necessary capabilities, and finally drive Agents to complete specific work through tasks.

Quick Start

For first-time use, it is recommended to follow this path to complete the setup:

  • Go to the Agent list and create a new Agent.
  • Fill in the Agent name, description, availability scope, identity definition, and behavioral boundaries.
  • Follow the installation guide to deploy the Agent service on the target host or virtual machine.
  • Prepare team-available Skills and MCP services in the global configuration.
  • Return to the specific Agent's workspace and only enable the Skills and MCP services needed by that Agent.
  • If collaboration is needed on external messaging platforms, connect a Channel for the Agent.
  • Create the first task, supplement the objective and context, and let the Agent start execution.
  • Create scheduled tasks for fixed-period work and view the results of each execution in the execution records.

Feature Entries

Feedback

Is this page helpful? ×