Omilia MCP Tools
OfficialThe Omilia MCP Tools server enables LLM clients to build, deploy, and operate conversational AI agents on the Omilia Cloud Platform (OCP) through natural language commands.
Discovery & Search: List and search OCP groups, agents, knowledge bases, Pathfinder projects, Orchestrator apps, miniapps, phone numbers, variable collections, and dialog logs.
Orchestrator Apps: Create, fetch, deploy, and test conversational flows; send messages to deployed applications to test agent replies.
Agent Management: Create and update Concierge and Task agents, manage instructions, configure sub-agents, and add escalation queues.
WebService Tools: Create and edit WebService miniapps, set welcome, initial, error, and reaction prompts, and attach tools to agents.
Knowledge Management: Create Pathfinder projects, crawl URLs into FAQs, and attach knowledge bases to agents.
Monitoring & Debugging: Retrieve full dialog logs for specific dialog IDs and search logs using filters such as app IDs, date ranges, caller phone numbers (ANI), and OCP group names.
Variables: Search variable collections and list all variables within a collection.
Integrates with Google's Gemini CLI as an MCP-compatible client, enabling users to interact with the Omilia Cloud Platform tools for conversational AI management.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Omilia MCP Tools search miniapps for customer support"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Omilia MCP Server
Documentation: learn.ocp.ai
A Model Context Protocol server that lets MCP-aware LLM clients build, deploy, and operate conversational AI agents on the Omilia Cloud Platform (OCP) through natural language. Create an orchestrator app, spin up an autonomous agent, attach a knowledge base from a URL, wire a webservice tool, deploy, and smoke-test the result — all from a chat prompt in your editor or desktop app.
The server runs locally and talks to OCP over authenticated HTTPS. No data leaves your machine except the requests your MCP client makes on your behalf.
Supported MCP clients
Claude Desktop, Claude Code, Cursor, VS Code (with Copilot), and Codex.
Related MCP server: MCPM CLI
Install
Recommended — interactive wizard (answers a few prompts, then confirms before writing anything):
npx github:omilia/mcp initThe wizard selects your client, collects credentials with masked input,
runs prereq checks (Node 20+, uv), installs the config, and prints a
PASS/FAIL smoke-test summary before exiting. See the
Interactive wizard section
in the installation guide for the full prompt sequence.
Non-interactive / CI — supply all flags to skip prompts:
npx github:omilia/mcp init --client <claude|claude-code|cursor|vscode|codex>See the full installation guide at docs/installation.md
for the .mcpb one-click bundle (Claude Desktop) and manual configuration
alternatives.
Authentication
Recommended: Keycloak password grant — set OCP_USERNAME, OCP_PASSWORD, and OCP_KEYCLOAK_REALM (default master).
Faster path: a Personal Access Token via OCP_ACCESS_TOKEN.
Both run in-memory only — credentials never persist. See docs/installation.md for the realm-by-environment table.
Tools (31)
The MCP surface, grouped by area. Each row is one tool your MCP-aware
LLM can invoke through tools/call.
Discovery & guides
Tool | Description |
| Return a markdown guide for a canonical OCP MCP workflow |
| List OCP groups the current user has access to |
| List all agents in scope |
| List knowledge bases (FAQ vector stores) |
| List Pathfinder projects (search filterable) |
| Search Orchestrator apps |
| Search miniapps |
| Search phone numbers attached to apps |
| Search variable collections |
| Search dialog logs by group / app / date |
Orchestrator apps
Tool | Description |
| Create a new Orchestrator app |
| Fetch an app's canvas |
| Wire a Concierge into an app's canvas |
| Deploy an Orchestrator app |
| Send a message to a deployed app and receive the agent's reply |
Agents (Concierge & Task)
Tool | Description |
| Create a Concierge or Task agent |
| Update an agent's instructions |
| Update the Concierge's sub-agent list |
| Add a sub-agent to a Concierge |
| Add escalation queue(s) to an agent |
WebService miniapps & agents
Tool | Description |
| Create a WebService miniapp (HTTP tool) |
| Fetch a miniapp's configuration |
| Edit a WebService miniapp's request config |
| Set welcome / initial / error prompts on a miniapp |
| Create a Task agent that uses a WebService miniapp |
| Add another WebService tool to a Task agent |
Pathfinder (knowledge)
Tool | Description |
| Create a Pathfinder project |
| Crawl a URL into a Pathfinder FAQ and attach it to an agent |
| Attach an existing knowledge base to an agent |
Variables & dialogs
Tool | Description |
| List variables in a collection |
| Fetch the full dialog log for a dialog ID |
Examples
End-to-end walkthroughs of the canonical journeys (build, deploy, test, extend) live at docs/examples/.
License
MIT. See LICENSE.
Provenance
This repository is auto-generated from an internal source on every
release. Generated from internal commit 333d04b1f44b7f54b0803eba5c1631dd8b101444 on branch
mirror/333d04b1f44b. This release exposes 31 tools
across src/main.py, src/server.py. PRs against this repository cannot be merged
back upstream — open issues for visibility and feedback; bug fixes are
tracked in the internal source.
Available Tools
10 toolsget_collection_variablesC
Get a list of all variables in a collection.
Args:
collection_id: The ID of the collection to get variables for
Inputs: collection_id (required).
get_dialog_logsB
Get the dialog logs for a specific dialog ID. Useful for retrieving conversation history and analytics.
Args:
dialog_id: The ID of the dialog to retrieve logs for
Returns:
The dialog log data as a dictionary
Inputs: dialog_id (required).
get_miniappC
Get a specific miniapp by its ID. Useful to return various information about a miniapp.
Args:
miniapp_id: The ID of the miniapp to retrieve
Returns:
The miniapp data as a dictionary
Inputs: miniapp_id (required).
get_orchestrator_appA
Get an Orchestrator application canvas by ID. Users can ask for this by saying "show me the app", "show me the canvas", "app contents" or "show me the flow". The resulting JSON is a graph structure of nodes and edges athat describes a dialog flow.
Args:
canvas_id: The ID of the canvas to get. This is the ID of the application canvas, contained in the search_orchestrator_apps results.
Inputs: canvas_id (required).
search_dialog_logsA
Search dialogs using various filter criteria. Can also be requested by users by saying "find sessions", "search logs" or "identify dialog logs"
Args:
apps (list): List of miniApp_ids or sandbox_flowapp_app_ids to filter by. This is not the same as the orchestrator app ID! One MUST get the sandbox_flowapp_app_id from the search_orchestrator_apps tool first.
from_date (str, optional): Start date/time in ISO format or milliseconds timestamp. Defaults to 24 hours ago.
to_date (str, optional): End date/time in ISO format or milliseconds timestamp. Defaults to now.
size (int, optional): Number of results to return. Defaults to 10
ani (list, optional): List of ANIs to filter by. ANI is the phone number of the caller.
dialog_group (str, optional): Dialog group ID to filter by
ocp_group_names (list, optional): List of OCP group names to filter by
region (str, optional): Region to filter by
application_layer (bool, optional): Whether to include application layer. Defaults to True
steps_gt (int, optional): Filter dialogs with steps greater than this number
Returns:
dict: Search results containing matching dialogs
Inputs: apps (required), from_date, to_date, size, +6 more.
search_miniappsC
Search miniapps. Useful to return a list of miniapps that match a search term. Args: search_term: Optional search term to filter miniapps
Inputs: search_term.
search_numbersC
Search (phone) numbers with optional search term.
Args:
search_term: Optional search term to filter numbers
Inputs: search_term.
search_orchestrator_appsC
Search Orchestrator apps with optional search term.
Args:
search_term: Optional search term to filter apps
Inputs: search_term.
search_variable_collectionsC
Search variable collections with optional search term.
Args:
search_term: Optional search term to filter variable collections
Inputs: search_term.
set_miniapp_promptB
Set various types of prompts for a specific miniapp. Unified interface for setting welcome, initial, error and reaction prompts.
Args:
miniapp_id: The ID of the miniapp to set the prompt for
prompt_type: The type of prompt to set. Can be one of:
- "welcome" - The welcome message prompt
- "initial" - The initial question prompt
- "error_no_interpretation" - When system cannot interpret the user's input
- "error_no_response" - When user provides no input
- "error_global_errors" - For general system errors
- "error_agent_request" - Response when user requests an agent
- "error_critical_error" - For critical system errors
- "error_max_disconfirmations" - When max confirmation retries reached
- "error_max_wrong_inputs" - When max invalid inputs reached
- "error_max_dtmf_inputs" - When max DTMF inputs reached
- "reaction_greeting" - Response to user greetings
- "reaction_no_match" - When input doesn't match expected responses
- "reaction_same_state" - When user repeats same input
- "reaction_nice_response" - Acknowledgement responses
prompt: The prompt text to set
Inputs: miniapp_id (required), prompt_type (required), prompt (required).
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 10 tools
Most tools have distinct purposes targeting different resources (collections, dialogs, miniapps, orchestrator apps, numbers), but some overlap exists: 'get_dialog_logs' and 'search_dialog_logs' both handle dialog logs, though one is for specific IDs and the other for filtered searches. The 'get_miniapp' and 'set_miniapp_prompt' tools are clearly differentiated by their actions. Overall, descriptions help clarify boundaries, but the dialog log tools could potentially cause confusion.
All tool names follow a consistent verb_noun pattern using snake_case, with verbs like 'get', 'search', and 'set' clearly indicating actions. The naming is highly predictable and readable throughout the set, with no deviations in style or convention.
With 10 tools, the count is well-scoped for managing a conversational AI platform, covering variables, dialogs, miniapps, orchestrator apps, and numbers. Each tool appears to earn its place by addressing specific operations without being overly sparse or bloated.
The toolset provides good read/search capabilities (get and search operations) but lacks update or delete functions for most resources, except for 'set_miniapp_prompt' which allows updates. There are notable gaps in lifecycle coverage: no tools for creating or deleting miniapps, orchestrator apps, variable collections, or numbers, which could limit agent workflows in managing these resources fully.
Related MCP Connectors
One OAuth connector for your team's remote MCP servers, with Tool Packs, guardrails and logs.
Your company's AI assistants, agents, workflows and knowledge bases from Orckai, as MCP tools.
Manage Nife.io cloud apps, sites, secrets and Git integrations from MCP clients.
Free MCP tools to validate multi-agent coordination thread turns and retrieve schema/samples.
Related MCP Servers
- AlicenseCqualityBmaintenanceMCP server for Voximplant API enabling calls, SMS, recordings, scenarios, and rules management with 11 tools and 2 skills.2386 npm3MIT
- AlicenseNot gradedqualityDmaintenanceA command-line tool for managing MCP servers in Claude App, which can also act as an MCP server itself to enable package management and discovery.35 npm106AGPL 3.0
- FlicenseAqualityDmaintenanceAn MCP server that provides tools to interact with CLI applications, enabling process management including start, stop, check status, send input, and read output.8-
- AlicenseAqualityDmaintenanceAn MCP server for querying Observe.com logs using OPAL, with tools for dataset discovery, schema inspection, and log searching.7MIT