Agent Capabilities
The OMD Cleo agent has access to a set of tools it can invoke during a conversation or workflow run. Tools give the LLM the ability to act on your data — not just generate text.
Available tools
| Tool group | What it does |
|---|---|
| OMD Go | Query tasks, trips, customers, resources and other OMD Go entities; replan tasks |
| OMD Powerhouse | Retrieve dashboards, charts, and run SQL queries against the Powerhouse BI data lake |
| Document Hub | Semantic search through documents stored in the OMD Document Hub |
| Memory | Store and retrieve facts about users and the organisation across conversations |
| Code Execution | Run Python code ad hoc or on a schedule in a managed, durable sandbox |
| Web Search | Search the web for current information |
| Email Send | Send emails on behalf of a user or workflow |
How tools work
During an agentic loop iteration, the LLM decides which tool to call based on the user's request. The agent:
- Calls the tool with the LLM-generated arguments.
- Appends the tool result to the conversation history.
- Calls the LLM again with the updated history.
- Repeats until the LLM produces a final response (no further tool calls).
Tool confirmation
For tools that modify data (e.g. updating a task, sending an email), the agent can be configured to request user confirmation before executing. The user approves or rejects the action via a message or reaction.
Tool availability per workflow
Not all tools are enabled for every workflow. The set of available tools is configured per workflow and per tenant. Tenant administrators can restrict or extend the default tool set.