OnSecurity MCP Server
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., "@OnSecurity MCP Server show me the top 10 most common findings across pentests"
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.
OnSecurity MCP
A Model Context Protocol (MCP) server for the OnSecurity API with the ability query rounds, findings, prerequisites, blocks and notifications.
Installation
Add the following to your Claude Desktop configuration file (adjust the paths as needed) and choose UAT or Prod:
{
"mcpServers": {
"onsec-mcp": {
"command": "npx",
"args": [
"-y",
"github:onsecurity/onsecurity-mcp-server"
],
"env": {
"ONSECURITY_API_TOKEN": "your_api_token",
"ONSECURITY_API_BASE": "https://app.onsecurity.io/api/v2"
}
}
}
}After adding this configuration, restart Claude Desktop, and you'll be able to access the OnSecurity tools through Claude.
Related MCP server: @konsulto/mcp
Usage
Once configured, Claude will have access to the following tools:
get-roundsget-findingsget-blocksget-notificationsget-prerequisites
Example Questions
Give me a summary of my most recent pentest/scan.
Show me trends across my pentests as a graph.
What can I address to make the most impact most quickly on my most recent pentest?
I would like summaries for different types of stakeholders on the state of our recent pentest engagemenets - eg high level, technical, managerial etc
Do I need to action anything to prevent test getting held up?
Are there any new findings?
What are the top 10 most common findings across pentests and scans.
Note: It is useful sometimes to configure Claude to "Extended thinking" for some questions.
Available Tools
5 toolsget-blocksB
Get all blocks data from OnSecurity. Blocks are reusable security finding templates that can be used across different assessments. They contain standardized vulnerability descriptions, risks, and recommendations. Note that you can get how often a block is used, which is a way to get the most common findings ACROSS ALL CLIENTS ONLY as blocks are the basis of findings across pentests and scans.
Inputs: round_type_id, approved, automation_approved, sort, +6 more.
get-findingsB
Get all findings data from OnSecurity from client in a high level summary, only include the summary, not the raw data and be sure to present the data in a way that is easy to understand for the client. You can optionally filter findings by round_id. HOWEVER ONLY USE THIS TOOL WHEN ASKED FOR FINDINGS RELATED TO A CLIENT OR MY FINDINGS, NOT THE BLOCKS TOOL.
Inputs: round_id, round_type, sort, limit, +5 more.
get-notificationsC
Get all notifications data from OnSecurity from client in a high level summary, only include the summary, not the raw data and be sure to present the data in a way that is easy to understand for the client.
Inputs: sort, limit, page, includes, +2 more.
get-prerequisitesC
Get all prerequisites data from OnSecurity for a specific round. Prerequisites are requirements that need to be fulfilled before a security assessment can begin.
Inputs: round_id (required), sort, limit, page, +2 more.
get-roundsC
Get all rounds data from OnSecurity from client in a high level summary. When replying, only include the summary, not the raw data and be sure to present the data in a way that is easy to understand for the client. Rounds can be pentest rounds, scan rounds, or radar rounds.
Inputs: round_type, sort, limit, page, +4 more.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 5 tools
Each tool has a clearly distinct purpose targeting different resources: blocks (security templates), findings (client security issues), notifications (client alerts), prerequisites (assessment requirements), and rounds (assessment cycles). The descriptions explicitly differentiate them, with notes like 'NOT THE BLOCKS TOOL' to prevent confusion, ensuring no overlap in functionality.
All tool names follow a consistent verb_noun pattern with hyphens (e.g., get-blocks, get-findings, get-notifications). This uniform naming convention makes the tool set predictable and easy to navigate, with no deviations in style or structure across the five tools.
With 5 tools, the count is reasonable for a security assessment server, covering key areas like templates, findings, notifications, prerequisites, and rounds. It is slightly lean but well-scoped; minor gaps like update or delete operations are acceptable given the server's focus on data retrieval for client summaries.
The tool set provides good read-only coverage for retrieving security assessment data, but there are notable gaps in CRUD operations. For example, there are no tools for creating, updating, or deleting findings, blocks, or rounds, which limits agents to viewing data without modifying it, potentially hindering workflow completeness in a security management context.
Related MCP Connectors
MCP server for the Vulners.com API
An MCP server that let you interact with Cycloid.io Internal Development Portal and Platform
An MCP server that provides read access to your cloud storage providers, bank accounts and more.
An MCP server that provides an API to LLMs to manage their JumpCloud resources.
Related MCP Servers
- AlicenseAqualityCmaintenanceMCP Server for the Notion API, enabling Claude to interact with Notion workspaces.313,089 npm925MIT

@konsulto/mcpofficial
AlicenseAqualityCmaintenanceMCP server that enables Claude Code to drive the Konsulto cybersecurity audit platform from the CLI, including reading and writing findings, managing evidence, and handling scope and assets.1934 npm1MIT
@repomend/mcpofficial
AlicenseNot gradedqualityDmaintenanceSecurity scanning MCP server that connects Claude to RepoMend findings, enabling vulnerability management and automated fix drafting.MIT- AlicenseAqualityDmaintenanceMCP server that connects Claude to Dependency-Track for natural language vulnerability triage, analysis, and management.14MIT