Dewain Robinson’s YouTube video walks viewers through the practical steps for modifying the manifest for Microsoft 365 Copilot agents. He explains both the low-code options available in Copilot Studio and the inline builder, and then shows how to edit manifest and config files directly in code for more advanced scenarios. Consequently, the video frames manifest edits as a way to unlock versioning and source control for agents, which matters for teams that need repeatable deployments and change tracking.
Manifests act as the configuration blueprint that tells Microsoft 365 and Teams how an agent should behave and appear. They include bot IDs, agent type, permissions, icons, and metadata, while a new copilot agents section makes custom engine agents visible inside Copilot chat. Therefore, editing the manifest is not just cosmetic; it changes registration, discovery, and capability exposure across Microsoft 365 experiences.
Robinson contrasts declarative agents, which rely on metadata-driven behavior, with custom engine agents, which give developers full orchestration control. He highlights that switching from declarative to custom engine agents involves manifest updates that explicitly declare the agent type and Bot ID. As a result, teams gain more control over workflows but also take on more responsibility for orchestration and security.
The video emphasizes recent changes that streamline how custom engine agents integrate with Copilot chat and Teams. In particular, support for a dedicated copilot agents manifest section helps agents register correctly so they appear in chat experiences, while tighter tooling in the ecosystem supports both low-code and code-driven approaches. Thus, organizations can choose a path that fits their developer skills and governance needs.
Additionally, Robinson covers the rollout of the Agents & Connectors dashboard in the Microsoft 365 Admin tools, which centralizes visibility and control. This dashboard lets admins filter agents, set sensitivity labels, and monitor performance, which supports compliance and risk management as agent usage grows. Consequently, manifest modification now ties directly into broader governance workflows rather than being an isolated developer task.
Choosing between declarative agents and custom engine agents creates clear tradeoffs: declarative agents are easier to maintain but offer limited orchestration control, while custom engine agents provide flexibility at the cost of complexity. Robinson points out that teams moving to custom engines must manage additional elements such as authentication, routing, and telemetry, which increases the operational burden. Therefore, organizations should weigh immediate developer convenience against long-term control and scalability.
Moreover, editing manifests directly introduces versioning and security questions that administrators must address. While source-controlling manifest files improves release management, it also requires stricter approval and review processes to prevent accidental exposure of sensitive settings. Consequently, balancing speed of change with governance and auditability becomes a central challenge for IT teams deploying Copilot agents.
Robinson recommends using tools like Visual Studio Code with the Microsoft 365 Agents Toolkit to edit manifests and validate changes before upload. He also underscores the value of checking manifest updates into source control so teams can track who changed what and when, which is essential for troubleshooting and compliance. Thus, following a predictable development workflow reduces risk and improves repeatability.
At the governance level, the video encourages administrators to adopt the new Agents & Connectors controls to assign sensitivity labels, manage connectors, and monitor agent health. In addition, teams should create clear policies for manifest change reviews and testing, and they should automate deployment where possible to reduce human error. By combining manifest-level discipline with centralized admin tools, organizations can scale agent deployments while keeping security and compliance under control.
In summary, Dewain Robinson’s video serves as a practical guide for teams looking to customize and control Microsoft 365 Copilot agents through manifest edits. It balances hands-on instructions with attention to governance, showing that deeper customization brings power but also responsibility. Therefore, teams should plan for both the technical steps and the organizational processes needed to manage manifest-driven agent deployments successfully.
M365 Copilot agent manifest, modify manifest M365 Copilot, edit Copilot agents manifest, update manifest Microsoft 365 Copilot, Copilot manifest schema, deploy Copilot agent manifest, Copilot manifest permissions, manifest troubleshooting M365 Copilot