Knuckles-Team/servicenow-api

ServiceNow - API | MCP | A2A | AG-UI

Version: 1.6.23

Overview

This project started out as a python wrapper for ServiceNow, but since the dawn of standards like Model Context Protocol (MCP) and Agent2Agent (A2A) Agent. This Agent has solved the issue with tons of MCP tools by distributing those tools to child agents. This allows you to run this entire agent on a small context window and maintain blazing speed! A 10K Context token minimum recommended, although I got away with 4K in some tests using qwen-4b on default settings.

This repository has become a sandbox for some brand new and pretty cool features with respects to each of those. The original APIs should remain stable, but please note, the latest features like A2A may still be unstable as it is under active development.

This project now includes an MCP server, which wraps all the original APIs you know and love in the base project. This allows any MCP capable LLM to leverage these tools and interact with ServiceNow. The MCP Server is enhanced with various authentication mechanisms, middleware for observability and control, and optional Eunomia authorization for policy-based access control.

ServiceNow A2A implements a multi-agent system designed to manage and interact with ServiceNow tasks through a delegated, specialist-based architecture. Built using Python, it leverages libraries like pydantic-ai for agent creation, FastMCPToolset for integrating Model Context Protocol (MCP) tools, and Graphiti for building a temporal knowledge graph from official ServiceNow documentation. The system runs as a FastAPI server via Uvicorn, exposing an Agent-to-Agent (A2A) interface for handling requests. pydantic-ai is able to expose your agent as an A2A agent out of the box with .to_a2a(). This allows this agent to be integrated in any agentic framework like Microsoft Agent Framework (MAF) or crew.ai. The core idea is a Supervisor Agent that acts as an intelligent router. It analyzes user queries and assigns tasks to specialized child agents, each focused on a specific domain (e.g., incidents, change management). Child agents have access to filtered MCP tools and a shared knowledge graph for reasoning over documentation, ensuring accurate and context-aware executions without direct access to the underlying APIs from the orchestrator.

This architecture is optimized for:

• Speed: Parallel execution potential and smaller context windows per agent.
• Context Usage: Each agent only loads the tools and skills relevant to its domain, allowing the system to handle hundreds of MCP tools without exceeding token limits or confusing the LLM.
• Modularity: New domains can be added as simple child agents without touching the core supervisor logic.

This architecture promotes modularity, scalability, and maintainability, making it suitable for enterprise integrations with platforms like ServiceNow.

Contributions are welcome!

API

API Calls

• Application Service
• Change Management
• CI/CD
• CMDB
• Import Sets
• Incident
• Knowledge Base
• Table
• Custom Endpoint

If your API call isn't supported, you can use the api_request tool to perform GET/POST/PUT/DELETE requests to any ServiceNow endpoint.

MCP

All the available API Calls above are wrapped in MCP Tools. You can find those below with their tool descriptions and associated tag.

MCP Tools

A2A Agent

Architecture Summary

The system follows a hierarchical multi-agent design:

• Orchestrator Agent: Acts as the entry point, parsing user requests and delegating to child agents based on domain tags.
• Child Agents: Domain-specific specialists (one per tag, e.g., "cmdb", "incidents") that execute tasks using filtered MCP tools and query the knowledge graph for guidance from official docs.
• Knowledge Graph (Graphiti): A centralized, temporal graph database ingesting ServiceNow documentation, accessible by child agents for retrieval-augmented reasoning.
• MCP Tools: Distributed via tags to ensure each child agent only accesses relevant tools, preventing overload and enhancing security.
• Server Layer: Exposes the orchestrator as an A2A API using pydantic-ai's to_a2a method and Uvicorn.

Key integrations:

• Pydantic-AI: Used for defining agents, models (e.g., OpenAI, Anthropic), and run contexts.
• A2A (Agent-to-Agent): Enables inter-agent communication through delegation tools.
• Graphiti: Builds a corpus from URLs of official docs, supporting backends like Kuzu (local), Neo4j, or FalkorDB.
• FastMCPToolset: Provides MCP wrappers for ServiceNow APIs, filtered by tags for distribution.

The script initializes the graph on startup (ingesting docs if needed), creates child agents with tag-filtered tools and graph access, builds the supervisor with delegation tools, and launches the server.

Architecture:

---
config:
  layout: dagre
---
flowchart TB
 subgraph subGraph0["Agent Capabilities"]
        D["MCP Tools"]
        F["Agent Skills"]
        C["Specialized Agents"]
  end
    A["User Query"] --> B["A2A Server - Uvicorn/FastAPI"]
    B --> G["Supervisor Agent"]
    G -- "Assigns Task" --> C
    C --> D & F
    D --> E["Platform API"]

     G:::supervisor
     C:::agent
     B:::server
     A:::server
    classDef server fill:#f9f,stroke:#333
    classDef agent fill:#bbf,stroke:#333,stroke-width:2px
    classDef supervisor fill:#ff9,stroke:#333,stroke-width:2px
    style B stroke:#000000,fill:#FFD600
    style D stroke:#000000,fill:#BBDEFB
    style F fill:#BBDEFB
    style A fill:#C8E6C9
    style subGraph0 fill:#FFF9C4

This diagram shows the flow from user input to delegation, tool execution, and knowledge retrieval. The orchestrator synthesizes results from children before responding.

Breakdown

1. Pydantic-AI Integration

Pydantic-AI is the backbone for agent modeling and execution. Agents are created using the Agent class, which takes an LLM model (e.g., OpenAIChatModel), system prompt, tools/toolsets, and a name.

• Model Creation: The create_model function supports multiple providers (OpenAI, Anthropic, Google, HuggingFace) via environment variables or CLI args.
• RunContext: Used in tool functions for contextual execution.
• Agent Hierarchy: Orchestrator has delegation tools (async functions wrapping child agent.run() calls). Children have MCP toolsets and Graphiti tools.

This enables structured, type-safe agent interactions with Pydantic validation.

2. A2A (Agent-to-Agent) Framework

A2A facilitates communication between agents:

• The orchestrator is converted to an A2A app via agent.to_a2a(), defining skills (high-level capabilities per tag).
• Delegation tools are async callables named delegate_to_{tag}, allowing the orchestrator to invoke children dynamically.
• This distributes workload, with the orchestrator synthesizing outputs for a cohesive response.

3. Multiple Agents

• Orchestrator: Analyzes queries to identify domains (tags), delegates via tools, and combines results. System prompt emphasizes delegation without direct action.
• Child Agents: One per tag in TAGS list (e.g., "incidents", "cmdb"). Each has a focused system prompt, filtered tools, and Graphiti access.
• Creation Flow: In create_orchestrator, children are instantiated in a loop, stored in a dict, and wrapped as tools for the parent.

This multi-agent setup scales by adding tags/children without refactoring the core.

4. Distribution of MCP Tools via Tags

MCP tools (from FastMCPToolset) wrap ServiceNow APIs. Distribution:

• Tools are tagged (e.g., "incidents" for incident-related endpoints).
• In create_child_agent, the toolset is filtered: filtered_toolset = toolset.filtered(lambda ctx, tool_def: tag in (tool_def.tags or [])).
• This ensures each child only sees relevant tools, reducing complexity and potential misuse.

Example: The "incidents" child gets tools like incident creation/query, while "cmdb" gets CMDB-specific ones.

5. Corpus of Knowledge from Official Documentation (Graphiti)

• Initialization: initialize_graphiti_db creates a Graphiti instance with the chosen backend (Kuzu for local, Neo4j/FalkorDB for servers).
• Ingestion: If the graph is empty or --graphiti-force-reinit is set, it fetches and adds episodes from INITIAL_DOC_URLS (ServiceNow API refs, guides).
• Temporal Aspect: Graphiti's episodes preserve dates for version-aware queries.
• Access by Children:
  • For Kuzu (embedded): Custom tools ingest_to_graph and query_graph wrap Graphiti methods.
  • For servers: Uses a separate FastMCPToolset for Graphiti MCP, filtered by tag.
• Usage: Children query the graph (e.g., "Retrieve details for Table API") to inform tool calls, enabling retrieval-augmented generation (RAG) over docs.

This creates a dynamic corpus, allowing agents to "understand" APIs without hardcoding.

Component Interaction Diagram

sequenceDiagram
    participant User
    participant Server as A2A Server
    participant Supervisor as Supervisor Agent
    participant Agent as Specialized Agent
    participant Skill as Agent Skills
    participant MCP as MCP Tools

    User->>Server: Send Query
    Server->>Supervisor: Invoke Supervisor
    Supervisor->>Agent: Assign Task (e.g. "Get Incident")
    Agent->>Skill: Analyze Skills Available
    Skill->>Agent: Provide Guidance on Next Steps
    Agent->>MCP: Invoke Tool
    MCP-->>Agent: Tool Response Returned
    Agent-->>Supervisor: Return Results Summarized
    Supervisor-->>Server: Final Response
    Server-->>User: Output

This sequence highlights delegation, knowledge retrieval, and tool execution.

Strengths

• Modularity & Scalability: Tags allow easy addition of domains (e.g., new ServiceNow modules) by extending TAGS and MCP tools. Graphiti scales with more docs via incremental ingestion.
• Knowledge-Driven Reasoning: By ingesting official docs, children can handle evolving APIs (e.g., Zurich bundle updates) without code changes—queries adapt to temporal data.
• Efficiency: Tool filtering prevents overload; delegation parallelizes tasks (though sequential in code, extensible to async).
• Flexibility: Supports multiple LLMs/backends via args/envs; A2A enables integration with other systems.
• Minimal Setup: Graph auto-initializes; defaults make it runnable out-of-box (assuming MCP/Graphiti servers).

Features:

• Authentication: Supports multiple authentication types including none (disabled), static (internal tokens), JWT, OAuth Proxy, OIDC Proxy, and Remote OAuth for external identity providers.
• Middleware: Includes logging, timing, rate limiting, and error handling for robust server operation.
• Eunomia Authorization: Optional policy-based authorization with embedded or remote Eunomia server integration.
• Resources: Provides instance_config and incident_categories for ServiceNow configuration and data.
• Prompts: Includes create_incident_prompt and query_table_prompt for AI-driven interactions.
• OIDC Token Delegation: Supports token exchange for ServiceNow API calls, enabling user-specific authentication via OIDC.
• OpenAPI JSON Tool Import: Import custom ServiceNow API Endpoints through the OpenAPI JSON generated.

Usage

API

OAuth Authentication

#!/usr/bin/python
# coding: utf-8
from servicenow_api.servicenow_api import Api

username = "<SERVICENOW USERNAME>"
password = "<SERVICENOW PASSWORD>"
client_id = "<SERVICENOW CLIENT_ID>"
client_secret = "<SERVICENOW_CLIENT_SECRET>"
servicenow_url = "<SERVICENOW_URL>"

client = Api(
    url=servicenow_url,
    username=username,
    password=password,
    client_id=client_id,
    client_secret=client_secret
)

table = client.get_table(table="<TABLE NAME>")
print(f"Table: {table.model_dump()}")

Basic Authentication

#!/usr/bin/python
# coding: utf-8
from servicenow_api.servicenow_api import Api

username = "<SERVICENOW USERNAME>"
password = "<SERVICENOW PASSWORD>"
servicenow_url = "<SERVICENOW_URL>"

client = Api(
    url=servicenow_url,
    username=username,
    password=password
)

table = client.get_table(table="<TABLE NAME>")
print(f"Table: {table.model_dump()}")

Proxy and SSL Verify

#!/usr/bin/python
# coding: utf-8
from servicenow_api.servicenow_api import Api

username = "<SERVICENOW USERNAME>"
password = "<SERVICENOW PASSWORD>"
servicenow_url = "<SERVICENOW_URL>"

proxies = {"https": "https://proxy.net"}

client = Api(
    url=servicenow_url,
    username=username,
    password=password,
    proxies=proxies,
    verify=False
)

table = client.get_table(table="<TABLE NAME>")
print(f"Table: {table.model_dump()}")

MCP

MCP CLI

Using as an MCP Server

The MCP Server can be run in two modes: stdio (for local testing) or http (for networked access). To start the server, use the following commands:

Run in stdio mode (default):

servicenow-mcp --transport "stdio"

Run in HTTP mode:

servicenow-mcp --transport "http"  --host "0.0.0.0"  --port "8000"

Run in Production:

Embedded Eunomia:

mcp_policies.json

{
  "policies": [
    {
      "id": "servicenow_read_policy",
      "description": "Allow read-only tools if user has read scope",
      "allow": true,
      "conditions": [
        {
          "tool": ["get_application", "get_cmdb", "batch_install_result"],
          "scopes": ["servicenow.read", "servicenow.full"]
        }
      ]
    },
    {
      "id": "servicenow_write_policy",
      "description": "Allow write tools if user has write scope and is admin",
      "allow": true,
      "conditions": [
        {
          "tool": ["batch_install", "batch_rollback", "app_repo_install"],
          "scopes": ["servicenow.write", "servicenow.full"],
          "claims": {"role": "admin"}
        }
      ]
    },
    {
      "id": "default_deny",
      "description": "Deny all other access",
      "allow": false
    }
  ]
}

Run command examples:

export IDENTITY_JWKS_URI="https://your-identity-provider.com/.well-known/jwks.json"
export API_IDENTIFIER="servicenow-mcp"
export PRODUCT_READ_SCOPE="mcpserverapi.product.read"
export INVENTORY_READ_SCOPE="mcpserverapi.inventory.read"
servicenow-mcp \
--transport "http"  \
--host "0.0.0.0" \
--port "8000" \
--auth-type "jwt" \
--token-jwks-uri "${IDENTITY_JWKS_URI}" \
--token-issuer "https://your-identity-provider.com" \
--token-audience "${API_IDENTIFIER}" \
--required-scopes "$PRODUCT_READ_SCOPE,$INVENTORY_READ_SCOPE" \
--eunomia-type "embedded" \
--eunomia-policy-file "mcp_policies.json"

# 1. JWKS (Production, RS256)
servicenow-mcp --auth-type jwt \
  --token-jwks-uri https://auth.example.com/.well-known/jwks.json \
  --token-issuer https://auth.example.com \
  --token-audience servicenow-mcp \
  --required-scopes servicenow.read,servicenow.write

# 2. HMAC (Internal, HS256)
servicenow-mcp --auth-type jwt \
  --token-secret "your-256-bit-secret-here-min-32-chars" \
  --token-algorithm HS256 \
  --token-issuer internal-auth \
  --token-audience mcp-api

# 3. Static RSA Key (Dev)
servicenow-mcp --auth-type jwt \
  --token-public-key ./public_key.pem \
  --token-issuer test-issuer \
  --token-audience test-mcp

# 4. With Delegation
--enable-delegation --auth-type jwt ... (uses JWT as subject_token)

#5 JWKS (Production, Asymmetric RS256)
servicenow-mcp --transport "http" --auth-type "jwt" \
  --token-jwks-uri "https://auth.example.com/.well-known/jwks.json" \
  --token-issuer "https://auth.example.com" \
  --token-audience "servicenow-mcp" \
  --required-scopes "servicenow.read,servicenow.write"

#6 HMAC (Internal/Microservices, HS256)
servicenow-mcp --transport "http" --auth-type "jwt" \
  --token-secret "your-256-bit-secret-min-32-chars" \
  --token-algorithm "HS256" \
  --token-issuer "internal-auth" \
  --token-audience "mcp-api"

#7 Static Public Key (Dev/Testing, RS256)
servicenow-mcp --transport "http" --auth-type "jwt" \
  --token-public-key "/path/to/public_key.pem" \
  --token-issuer "test-issuer" \
  --token-audience "test-mcp"

Native Fast MCP Arguments

# Enable JWT verification
export FASTMCP_SERVER_AUTH=fastmcp.server.auth.providers.jwt.JWTVerifier

# For asymmetric verification with JWKS endpoint:
export FASTMCP_SERVER_AUTH_JWT_JWKS_URI="https://auth.company.com/.well-known/jwks.json"
export FASTMCP_SERVER_AUTH_JWT_ISSUER="https://auth.company.com"
export FASTMCP_SERVER_AUTH_JWT_AUDIENCE="mcp-production-api"
export FASTMCP_SERVER_AUTH_JWT_REQUIRED_SCOPES="read:data,write:data"

# OR for symmetric key verification (HMAC):
export FASTMCP_SERVER_AUTH_JWT_PUBLIC_KEY="your-shared-secret-key-minimum-32-chars"
export FASTMCP_SERVER_AUTH_JWT_ALGORITHM="HS256"  # or HS384, HS512
export FASTMCP_SERVER_AUTH_JWT_ISSUER="internal-auth-service"
export FASTMCP_SERVER_AUTH_JWT_AUDIENCE="mcp-internal-api"

Deploy MCP Server as a Service:

Note: The ServiceNow MCP server can be deployed using Docker, with configurable authentication, middleware, and Eunomia authorization.

Docker Run:

docker pull knucklessg1/servicenow:latest

docker run -d \
  --name servicenow-mcp \
  -p 8004:8004 \
  -e HOST=0.0.0.0 \
  -e PORT=8004 \
  -e TRANSPORT=http \
  -e AUTH_TYPE=none \
  -e EUNOMIA_TYPE=none \
  -e SERVICENOW_INSTANCE=https://yourinstance.servicenow.com \
  -e SERVICENOW_USERNAME=user \
  -e SERVICENOW_PASSWORD=pass \
  -e SERVICENOW_CLIENT_ID=client_id \
  -e SERVICENOW_CLIENT_SECRET=client_secret \
  -e SERVICENOW_VERIFY=False \
  knucklessg1/servicenow:latest

For advanced authentication (e.g., OIDC Proxy with token delegation) or Eunomia, add the relevant environment variables:

For Additional OpenAPI Tool Import, include OPENAPI_FILE.

docker run -d \
  --name servicenow-mcp \
  -p 8004:8004 \
  -e HOST=0.0.0.0 \
  -e PORT=8004 \
  -e TRANSPORT=http \
  -e AUTH_TYPE=oidc-proxy \
  -e FASTMCP_SERVER_AUTH_JWT_ALGORITHM=HS256 \
  -e FASTMCP_SERVER_AUTH_JWT_PUBLIC_KEY="your-shared-secret" \
  -e FASTMCP_SERVER_AUTH_JWT_REQUIRED_SCOPES="servicenow.read,servicenow.write" \
  -e OIDC_CONFIG_URL=https://provider.com/.well-known/openid-configuration \
  -e OIDC_CLIENT_ID=your-client-id \
  -e OIDC_CLIENT_SECRET=your-client-secret \
  -e OIDC_BASE_URL=https://your-server.com \
  -e ALLOWED_CLIENT_REDIRECT_URIS=http://localhost:*,https://*.example.com/* \
  -e ENABLE_DELEGATION=True \
  -e SERVICENOW_AUDIENCE=https://yourinstance.servicenow.com \
  -e DELEGATED_SCOPES="api user_impersonation" \
  -e EUNOMIA_TYPE=embedded \
  -e EUNOMIA_POLICY_FILE=/app/mcp_policies.json \
  -e SERVICENOW_INSTANCE=https://yourinstance.servicenow.com \
  -e SERVICENOW_USERNAME=user \
  -e SERVICENOW_PASSWORD=pass \
  -e SERVICENOW_CLIENT_ID=client_id \
  -e SERVICENOW_CLIENT_SECRET=client_secret \
  -e SERVICENOW_VERIFY=False \
  -e OPENAPI_FILE=/app/servicenow_openapi.json \
  knucklessg1/servicenow:latest

Configure mcp.json for AI Integration

Recommended: Store secrets in environment variables with lookup in the JSON file.

For Testing Only: Plain text storage will also work, although not recommended.

{
  "mcpServers": {
    "servicenow": {
      "command": "uv",
      "args": [
        "run",
        "--with",
        "servicenow-api",
        "servicenow-mcp",
        "--transport",
        "${TRANSPORT}",
        "--host",
        "${HOST}",
        "--port",
        "${PORT}",
        "--auth-type",
        "${AUTH_TYPE}",
        "--eunomia-type",
        "${EUNOMIA_TYPE}",
        "--enable-delegation",
        "${ENABLE_DELEGATION}",
        "--servicenow-audience",
        "${SERVICENOW_AUDIENCE}",
        "--delegated-scopes",
        "${DELEGATED_SCOPES}"
      ],
      "env": {
        "SERVICENOW_INSTANCE": "https://yourinstance.servicenow.com",
        "SERVICENOW_USERNAME": "user",
        "SERVICENOW_PASSWORD": "pass",
        "SERVICENOW_CLIENT_ID": "client_id",
        "SERVICENOW_CLIENT_SECRET": "client_secret",
        "SERVICENOW_VERIFY": "False",
        "TOKEN_JWKS_URI": "${TOKEN_JWKS_URI}",
        "TOKEN_ISSUER": "${TOKEN_ISSUER}",
        "TOKEN_AUDIENCE": "${TOKEN_AUDIENCE}",
        "OAUTH_UPSTREAM_AUTH_ENDPOINT": "${OAUTH_UPSTREAM_AUTH_ENDPOINT}",
        "OAUTH_UPSTREAM_TOKEN_ENDPOINT": "${OAUTH_UPSTREAM_TOKEN_ENDPOINT}",
        "OAUTH_UPSTREAM_CLIENT_ID": "${OAUTH_UPSTREAM_CLIENT_ID}",
        "OAUTH_UPSTREAM_CLIENT_SECRET": "${OAUTH_UPSTREAM_CLIENT_SECRET}",
        "OAUTH_BASE_URL": "${OAUTH_BASE_URL}",
        "OIDC_CONFIG_URL": "${OIDC_CONFIG_URL}",
        "OIDC_CLIENT_ID": "${OIDC_CLIENT_ID}",
        "OIDC_CLIENT_SECRET": "${OIDC_CLIENT_SECRET}",
        "OIDC_BASE_URL": "${OIDC_BASE_URL}",
        "REMOTE_AUTH_SERVERS": "${REMOTE_AUTH_SERVERS}",
        "REMOTE_BASE_URL": "${REMOTE_BASE_URL}",
        "ALLOWED_CLIENT_REDIRECT_URIS": "${ALLOWED_CLIENT_REDIRECT_URIS}",
        "EUNOMIA_TYPE": "${EUNOMIA_TYPE}",
        "EUNOMIA_POLICY_FILE": "${EUNOMIA_POLICY_FILE}",
        "EUNOMIA_REMOTE_URL": "${EUNOMIA_REMOTE_URL}",
        "ENABLE_DELEGATION": "${ENABLE_DELEGATION}",
        "SERVICENOW_AUDIENCE": "${SERVICENOW_AUDIENCE}",
        "DELEGATED_SCOPES": "${DELEGATED_SCOPES}"
      },
      "timeout": 200000
    }
  }
}

Middleware

The MCP server includes the following built-in middleware for enhanced functionality:

• ErrorHandlingMiddleware: Provides comprehensive error logging and transformation.
• RateLimitingMiddleware: Limits request frequency with a token bucket algorithm (10 requests/second, burst capacity of 20).
• TimingMiddleware: Tracks execution time of requests.
• LoggingMiddleware: Logs all requests and responses for observability.
• UserTokenMiddleware: Extracts Bearer tokens for OIDC token delegation to ServiceNow (enabled with --enable-delegation).

Eunomia Authorization

The server supports optional Eunomia authorization for policy-based access control:

• Disabled (none): No authorization checks.
• Embedded (embedded): Runs an embedded Eunomia server with a local policy file (mcp_policies.json by default).
• Remote (remote): Connects to an external Eunomia server for centralized policy decisions.

To configure Eunomia policies:

Embedded Eunomia:

mcp_policies.json

{
  "policies": [
    {
      "id": "servicenow_read_policy",
      "description": "Allow read-only tools if user has read scope",
      "allow": true,
      "conditions": [
        {
          "tool": ["get_application", "get_cmdb", "batch_install_result"],
          "scopes": ["servicenow.read", "servicenow.full"]
        }
      ]
    },
    {
      "id": "servicenow_write_policy",
      "description": "Allow write tools if user has write scope and is admin",
      "allow": true,
      "conditions": [
        {
          "tool": [
            "batch_install",
            "batch_rollback",
            "app_repo_install"
          ],
          "scopes": [
            "servicenow.write",
            "servicenow.full"
          ],
          "claims": {
            "role": "admin"
          }
        }
      ]
    },
    {
      "id": "default_deny",
      "description": "Deny all other access",
      "allow": false
    }
  ]
}

On a dedicated server, run the following to initialize eunomia-mcp.

# Initialize a default policy file
eunomia-mcp init

# Validate the policy file
eunomia-mcp validate mcp_policies.json

A2A CLI

Endpoints

• Web UI: http://localhost:8000/ (if enabled)
• A2A: http://localhost:8000/a2a (Discovery: /a2a/.well-known/agent.json)
• AG-UI: http://localhost:8000/ag-ui (POST)

Using as an A2A Agent

The servicenow-agent tool hosts an Agent-to-Agent (A2A) server that orchestrates specialized child agents for different ServiceNow domains.

CLI mode:

servicenow-agent --port 9000 --provider openai --model-id gpt-4o

Docker:

docker run -d \
  -p 9000:9000 \
  -e SERVICENOW_INSTANCE=https://yourinstance.servicenow.com \
  -e SERVICENOW_USERNAME=user \
  -e SERVICENOW_PASSWORD=pass \
  knucklessg1/servicenow:latest \
  servicenow-agent \
  --port 9000 \
  --mcp-url http://host.docker.internal:8000/mcp

Note: The A2A agent requires a running MCP server to function.

Changelog

v1.4.7

• Multi-Agent Architecture: Introduced a Supervisor Agent to orchestrate tasks across specialized domain agents (Application, CMDB, Incidents, etc.).
• Supervisor Routing: The supervisor now intelligently "assigns" tasks rather than just delegating, using a clearer tool set.
• Tag-Based Tool Partitioning: MCP tools are now strictly partitioned by tags, ensuring agents only access relevant tools. This optimization allows for handling a massive number of tools without context overflow.
• Robust Tool Filtering: Enhanced name prefix handling in filters to ensuring tools like servicenow_get_incidents are correctly routed to the incidents agent.
• Smart Incident Lookup: get_incidents now automatically detects if an input is a number (INC...) vs a sys_id and switches query modes accordingly.
• Cache Busting: Improved Docker reliability with cache-busting mechanisms.

You can run the A2A server using the same Docker image by overriding the command.

Compose

Refer to the compose file within the repository for a great example how to get started with a fully local setup!

Run the service:

docker compose up -d

or

podman compose up -d

Install Python Package

python -m pip install servicenow-api[mcp,a2a,all]

Tests

python ./test/test_servicenow_models.py

Repository Owners