Skip to main content
Glama
DefectDojo MCP Server

DefectDojo MCP Server

by dcianciulli

DefectDojo MCP Server

A Model Context Protocol (MCP) server that provides comprehensive access to DefectDojo vulnerability management platform, plus a portable Agent Skill that teaches any coding agent how to use it correctly.

Repo: dcianciulli/defectdojo-mcp-server. The server can be run straight from this repository (no clone needed).

What's in this repository

Path

Purpose

src/defectdojo_mcp/

The MCP server (78 tools, Python/FastMCP, stdio transport)

skills/defectdojo/SKILL.md

Portable Agent Skill (standard SKILL.md format) — works with Claude Code, OpenCode, Codex CLI, Cursor, Gemini CLI and any other skill-compatible harness

skills/README.md

Install & distribution guide: per-harness install paths and MCP config snippets

.claude-plugin/

Claude Code plugin + marketplace manifests (bundles skill + MCP server in one install)

tests/

Live integration tests (adapt to your instance: point DEFECTDOJO_URL at a test instance)

Skill = knowledge (procedures, workflows, tool map — plain markdown, portable everywhere). MCP server = execution (the actual tools). They are independent: install either or both. Note: a skill alone does not provision MCP servers — that stays explicit host configuration (except via the Claude Code plugin below, which bundles both with user approval).

Related MCP server: CIRCL CVE SEARCH MCP Server

Quick start

Prerequisites: uv installed; environment variables DEFECTDOJO_URL (e.g. https://your-defectdojo.example.com) and DEFECTDOJO_API_KEY (token from DefectDojo → your profile → API v2 Key).

Install the skill (Node.js; auto-detects installed agents — Claude Code, OpenCode, Codex, Cursor…):

npx skills add https://github.com/dcianciulli/defectdojo-mcp-server -g

Run the MCP server (most harnesses — Kiro, Claude Desktop, Cursor, Gemini CLI share this JSON shape):

{
  "mcpServers": {
    "defectdojo": {
      "command": "uvx",
      "args": ["--from", "git+https://github.com/dcianciulli/defectdojo-mcp-server", "defectdojo-mcp-server"],
      "env": { "DEFECTDOJO_URL": "https://your-defectdojo.example.com", "DEFECTDOJO_API_KEY": "your-token" }
    }
  }
}

Other harnesses (OpenCode, Codex CLI, Gemini CLI, Kiro) use different config files and field names: exact snippets for each are in skills/README.md.

Claude Code one-step install (skill + MCP server together, via plugin):

claude plugin marketplace add dcianciulli/defectdojo-mcp-server
claude plugin install defectdojo@defectdojo-mcp

The plugin bundles the MCP server definition (runs via uvx straight from this repo); you still need the two environment variables above set in your environment.

Features

  • Findings: search with extensive filters (severity, status, SLA, CVE, reporter…), create, update, verify, duplicates

  • Finding lifecycle: close as false positive, mitigated, or duplicate — with closure note; reopen/reactivate

  • Risk acceptance: create with mandatory expiration date (findings auto-reactivate on expiry), update, expire early, reinstate

  • Products/Assets: CRUD with filtering by organization, lifecycle, tags (v3 API: assets)

  • Engagements & Tests: full lifecycle, notes

  • Scan Import/Reimport: results from 180+ scanners

  • Notes & metadata: add/remove finding notes, note types, metadata

  • Organizations, Users, Endpoints/Locations, JIRA integration, System administration (settings, Celery, SLA, notifications)

Finding lifecycle & risk acceptance (the important part)

Closing a finding requires a reason — use the dedicated tool, never raw status flips:

Reason

Tool

Not a real vulnerability

close_finding_false_positive(finding_id, note) (+ out_of_scope=true if applicable)

Vulnerability remediated

close_finding_mitigated(finding_id, note)

Duplicate of another finding

close_finding_duplicate(finding_id, duplicate_of)

Undo any closure with reopen_finding.

Risk acceptance is deliberately discouraged in favor of fixing: accept_risk requires a future expiration_date (rejected otherwise), an accepted_by and a justification; findings are reactivated automatically when the acceptance expires (reactivate_expired=true by default). Manage acceptances with update_risk_acceptance, expire_risk_acceptance, reinstate_risk_acceptance.

When asked to work on a project, resolve it by asset name first (list_products(name=…)); ask the user when ambiguous; use the organization only as a secondary hint. The bundled skill enforces all of this — installing it is recommended.

Available tools (78)

Findings (8)

list_findings · get_finding · create_finding · update_finding · verify_finding · delete_finding · get_finding_duplicates · reset_finding_duplicate_status

Finding lifecycle (5)

close_finding (generic, closure_type: false_positive / mitigated / duplicate) · close_finding_false_positive · close_finding_mitigated · close_finding_duplicate · reopen_finding

Risk acceptance (9)

accept_risk (expiration mandatory) · accept_risks_vulnerability (bulk by CVE) · list_risk_acceptances · get_risk_acceptance · create_risk_acceptance · update_risk_acceptance · delete_risk_acceptance · expire_risk_acceptance · reinstate_risk_acceptance

Notes & metadata (6)

list_finding_notes · add_finding_note · remove_finding_note · list_note_types · list_finding_metadata · add_finding_metadata

Products/Assets (5)

list_products / get_product / create_product / update_product / delete_product

Organizations (5)

list_organizations / get_organization / create_organization / update_organization / delete_organization

Engagements (8)

list_engagements / get_engagement / create_engagement / update_engagement · close_engagement / delete_engagement · list_engagement_notes / add_engagement_note

Tests (8)

list_tests / get_test / create_test / update_test / delete_test · list_test_types · list_test_notes / add_test_note

Scan import (2)

import_scan · reimport_scan

Endpoints (3)

list_endpoints · get_endpoint · list_endpoint_status

JIRA (5)

list_jira_instances / get_jira_instance / create_jira_instance · list_jira_projects / list_jira_finding_mappings

Users (6)

list_users / get_user / get_current_user / create_user / update_user / delete_user

System (8)

get_system_settings / update_system_settings · get_celery_status / get_celery_queue_details / purge_celery_queue / purge_celery_task · list_notifications / list_sla_configurations

Configuration

Environment variables

Variable

Required

Description

DEFECTDOJO_URL

Yes

Base URL of the DefectDojo instance (e.g. https://your-defectdojo.example.com)

DEFECTDOJO_API_KEY

Yes

API token (DefectDojo → user profile → API v2 Key)

Getting your API key

  1. Log into DefectDojo

  2. Open your profile (top-right menu → API v2 Key)

  3. Copy the token value

Authentication model

The token inherits the permissions of the user who generated it; every action (closures, notes, risk acceptances) is attributed to that user in DefectDojo. Each person should use their own token — do not share tokens between team members.

Supported scan types

DefectDojo supports 180+ scan types (ZAP, Burp, Nessus, Qualys, SARIF, Trivy, Grype, SonarQube, Semgrep, Dependency Check, Snyk, AWS Security Hub/Prowler, GitLab SAST/DAST, Checkmarx, Veracode…). Run list_test_types to get the full list from your instance.

Development

git clone <repo-url> && cd defectdojo-mcp-server
uv sync                 # creates .venv from pyproject.toml + uv.lock
uv run defectdojo-mcp-server          # start server locally (stdio)
  • Requires Python >= 3.10. The mcp dependency is pinned to 1.x (FastMCP API); uv.lock is committed to keep it that way.

  • tests/test_lifecycle_live.py exercises the full lifecycle (close FP/mitigated, reopen, risk acceptance with expiration, expire) against a live DefectDojo instance — adapt the finding ID and run it only against a test finding you own (never production data).

  • After changing the code, restart the MCP connection in your harness to pick up the new tools.

License

MIT

Available Tools

78 tools
accept_riskA

Accept the risk of one or more findings (risk acceptance with mandatory expiration date).

Creates a RiskAcceptance object. The expiration date is MANDATORY: findings are reactivated automatically when it passes (unless reactivate_expired=False). Risk acceptance is discouraged in favor of fixing; always require an explicit business justification and a named approver.

Args: finding_ids: List of finding IDs to accept accepted_by: Name/email of the person accepting the risk justification: Business justification for accepting expiration_date: Mandatory expiration (ISO date "YYYY-MM-DD" or full ISO datetime). Must be in the future. decision: Risk treatment by risk owner: A=Accept, V=Avoid, M=Mitigate, F=Fix, T=Transfer decision_details: Details of the decision / compensating controls recommendation: Security team recommendation: A=Accept, V=Avoid, M=Mitigate, F=Fix, T=Transfer recommendation_details: Details of the recommendation reactivate_expired: Reactivate findings when the acceptance expires (default True) restart_sla_expired: Restart SLA when the acceptance expires (default False) owner_id: Owner user ID (defaults to the authenticated user) name: Acceptance name (defaults to "Risk acceptance ( findings)")

Inputs: accepted_by (required), finding_ids (required), justification (required), expiration_date (required), +8 more. Returns: object

accept_risks_vulnerabilityA

Bulk accept risk for findings matching vulnerability IDs (CVEs).

Uses the native /findings/accept_risks/ endpoint: matches findings by vulnerability_id (CVE) inside the CURRENT engagement context. Use accept_risk() instead when you need explicit finding IDs and an expiration date.

Args: vulnerability_ids: CVE or advisory IDs (e.g. ["CVE-2024-1234"]) accepted_by: Name/email of the person accepting the risk justification: Justification for accepting findings with these vulnerability IDs

Inputs: accepted_by (required), justification (required), vulnerability_ids (required). Returns: object

add_engagement_noteC

Add a note to an engagement.

Args: engagement_id: The engagement ID entry: Note text content private: Whether the note is private

Inputs: entry (required), engagement_id (required), private. Returns: object

add_finding_metadataC

Add metadata to a finding.

Args: finding_id: The finding ID name: Metadata key name value: Metadata value

Inputs: name (required), value (required), finding_id (required). Returns: object

add_finding_noteB

Add a note to a finding.

Args: finding_id: The finding ID entry: Note text content private: Whether the note is private (visible only to you and superusers) note_type: Note type ID (see list_note_types)

Inputs: entry (required), finding_id (required), private, note_type. Returns: object

add_test_noteC

Add a note to a test.

Args: test_id: The test ID entry: Note text content private: Whether the note is private

Inputs: entry (required), test_id (required), private. Returns: object

close_engagementC

Close an engagement.

Args: engagement_id: The engagement ID to close

Inputs: engagement_id (required). Returns: object

close_findingA

Close a finding. Requires closure_type; use the dedicated tools below when unsure.

Prefer the semantic aliases:

  • close_finding_false_positive(finding_id, note)

  • close_finding_mitigated(finding_id, note)

  • close_finding_duplicate(finding_id, duplicate_of)

Args: finding_id: The finding ID to close closure_type: One of "false_positive", "mitigated", "duplicate" note: Optional note recorded on the finding at closure time note_type: Optional note type ID (see list_note_types) mitigated: Mitigation datetime (ISO 8601, defaults to now) out_of_scope: Mark finding as out of scope (only meaningful with closure_type="false_positive") duplicate: Mark finding as duplicate (only meaningful with closure_type="duplicate")

Inputs: finding_id (required), note, duplicate, mitigated, +3 more. Returns: object

close_finding_duplicateC

Close a finding as duplicate of another finding.

Args: finding_id: The finding ID to close duplicate_of: The ID of the original (canonical) finding note: Optional note

Inputs: finding_id (required), duplicate_of (required), note. Returns: object

close_finding_false_positiveB

Close a finding as a false positive (and optionally out of scope).

Args: finding_id: The finding ID to close note: Why this is a false positive (recommended) note_type: Optional note type ID (see list_note_types) out_of_scope: Also mark the finding as out of scope

Inputs: finding_id (required), note, note_type, out_of_scope. Returns: object

close_finding_mitigatedA

Close a finding because the vulnerability was remediated (fix applied).

Args: finding_id: The finding ID to close note: How/where it was fixed (recommended) note_type: Optional note type ID (see list_note_types) mitigated: Mitigation datetime (ISO 8601, defaults to now server-side)

Inputs: finding_id (required), note, mitigated, note_type. Returns: object

create_engagementC

Create a new engagement.

Args: name: Engagement name product_id: Product ID this engagement belongs to target_start: Target start date (YYYY-MM-DD) target_end: Target end date (YYYY-MM-DD) engagement_type: Type (Interactive or CI/CD) status: Status (Not Started, In Progress, Completed) description: Description lead_id: Lead user ID build_id: Build identifier commit_hash: Commit hash branch_tag: Branch or tag name source_code_management_uri: SCM URI deduplication_on_engagement: Dedupe scoped to engagement tags: List of tags

Inputs: name (required), product_id (required), target_end (required), target_start (required), +10 more. Returns: object

create_findingC

Create a new finding.

Args: title: Finding title severity: Severity level (Critical, High, Medium, Low, Info) test_id: ID of the test this finding belongs to description: Detailed description mitigation: Recommended mitigation impact: Impact description steps_to_reproduce: Steps to reproduce references: External references cwe: CWE identifier number active: Whether finding is active (default True) verified: Whether finding is verified (default False) numerical_severity: Numeric severity (S0-S4) line: Source code line number file_path: Source file path component_name: Affected component component_version: Component version static_finding: Is static analysis finding dynamic_finding: Is dynamic analysis finding tags: List of tags

Inputs: title (required), test_id (required), severity (required), cwe, +15 more. Returns: object

create_jira_instanceB

Create a new JIRA instance configuration.

Args: url: JIRA instance URL username: JIRA username password: JIRA API token or password default_issue_type: Default issue type for new issues epic_name_id: Custom field ID for epic name open_status_key: Transition ID for opening issues close_status_key: Transition ID for closing issues info_mapping_severity: JIRA priority for Info severity low_mapping_severity: JIRA priority for Low severity medium_mapping_severity: JIRA priority for Medium severity high_mapping_severity: JIRA priority for High severity critical_mapping_severity: JIRA priority for Critical severity

Inputs: url (required), password (required), username (required), epic_name_id, +8 more. Returns: object

create_organizationC

Create a new organization (product type).

Args: name: Organization name description: Description critical_product: Is critical product flag key_product: Is key product flag

Inputs: name (required), description, key_product, critical_product. Returns: object

create_productC

Create a new product (asset).

Args: name: Product name organization_id: Organization (product type) ID description: Product description lifecycle: Lifecycle stage (construction, production, retirement) platform: Platform (web service, desktop, iot, mobile, web) origin: Origin (third party library, purchased, contractor, internal, open source, outsourced) business_criticality: Business criticality (very high, high, medium, low, very low, none) external_audience: Has external audience internet_accessible: Is internet accessible tags: List of tags

Inputs: name (required), organization_id (required), tags, origin, +6 more. Returns: object

create_risk_acceptanceA

Create a new risk acceptance (low-level: prefer accept_risk, which enforces expiration).

Args: name: Risk acceptance name owner_id: Owner user ID accepted_findings: List of finding IDs to accept recommendation: Security recommendation code (A, V, M, F, T) recommendation_details: Details about the recommendation decision: Risk treatment decision code (A, V, M, F, T) decision_details: Details about the decision / compensating controls expiration_date: Expiration date (YYYY-MM-DD or ISO datetime; mandatory in accept_risk) accepted_by: Name/email of the person accepting the risk reactivate_expired: Reactivate findings when the acceptance expires (default True) restart_sla_expired: Restart SLA when the acceptance expires (default False)

Inputs: name (required), owner_id (required), accepted_findings (required), decision, +7 more. Returns: object

create_testC

Create a new test.

Args: engagement_id: Engagement ID this test belongs to test_type_id: Test type ID target_start: Target start datetime (YYYY-MM-DD or ISO format) target_end: Target end datetime title: Test title description: Test description lead_id: Lead user ID environment_id: Environment ID version: Version string branch_tag: Branch or tag build_id: Build identifier commit_hash: Commit hash tags: List of tags

Inputs: target_end (required), target_start (required), test_type_id (required), engagement_id (required), +9 more. Returns: object

create_userC

Create a new user.

Args: username: Username first_name: First name last_name: Last name email: Email address is_active: Active status is_superuser: Superuser status

Inputs: username (required), email, is_active, last_name, +2 more. Returns: object

delete_engagementC

Delete an engagement.

Args: engagement_id: The engagement ID to delete

Inputs: engagement_id (required). Returns: object

delete_findingC

Delete a finding.

Args: finding_id: The finding ID to delete

Inputs: finding_id (required). Returns: object

delete_organizationC

Delete an organization.

Args: organization_id: The organization ID to delete

Inputs: organization_id (required). Returns: object

delete_productC

Delete a product.

Args: product_id: The product ID to delete

Inputs: product_id (required). Returns: object

delete_risk_acceptanceC

Delete a risk acceptance.

Args: risk_acceptance_id: The risk acceptance ID

Inputs: risk_acceptance_id (required). Returns: object

delete_testC

Delete a test.

Args: test_id: The test ID to delete

Inputs: test_id (required). Returns: object

delete_userC

Delete a user.

Args: user_id: The user ID to delete

Inputs: user_id (required). Returns: object

expire_risk_acceptanceA

Expire a risk acceptance early (reactivates its findings).

Args: risk_acceptance_id: The risk acceptance ID reason: Optional reason for expiring

Inputs: risk_acceptance_id (required), reason. Returns: object

get_celery_queue_detailsA

Get per-task breakdown of the Celery queue.

Returns task names, counts, and queue positions. May be slow for large queues.

Inputs: no named parameters. Returns: result

get_celery_statusA

Get Celery worker and queue status.

Returns worker liveness, pending queue length, and configuration.

Inputs: no named parameters. Returns: object

get_current_userA

Get the currently authenticated user's profile.

Inputs: no named parameters. Returns: object

get_endpointB

Get a single endpoint by ID.

Args: endpoint_id: The endpoint ID

Inputs: endpoint_id (required). Returns: object

get_engagementC

Get a single engagement by ID.

Args: engagement_id: The engagement ID

Inputs: engagement_id (required). Returns: object

get_findingC

Get a single finding by ID.

Args: finding_id: The finding ID

Inputs: finding_id (required). Returns: object

get_finding_duplicatesC

Get duplicate findings for a given finding.

Args: finding_id: The finding ID

Inputs: finding_id (required). Returns: object

get_jira_instanceC

Get a JIRA instance configuration by ID.

Args: instance_id: The JIRA instance ID

Inputs: instance_id (required). Returns: object

get_organizationC

Get a single organization by ID.

Args: organization_id: The organization (product type) ID

Inputs: organization_id (required). Returns: object

get_productC

Get a single product by ID.

Args: product_id: The product ID

Inputs: product_id (required). Returns: object

get_risk_acceptanceB

Get a risk acceptance by ID.

Args: risk_acceptance_id: The risk acceptance ID

Inputs: risk_acceptance_id (required). Returns: object

get_system_settingsB

Get DefectDojo system settings.

Inputs: no named parameters. Returns: object

get_testC

Get a single test by ID.

Args: test_id: The test ID

Inputs: test_id (required). Returns: object

get_userC

Get a single user by ID.

Args: user_id: The user ID

Inputs: user_id (required). Returns: object

import_scanA

Import a scan report into DefectDojo.

Either engagement_id OR (product_name + auto_create_context=True) must be provided. Either file_path (local file) or scan_data (raw content) must be provided.

Args: scan_type: Scanner type (e.g., "ZAP Scan", "Nessus Scan", "SARIF", "Trivy Scan", etc.) file_path: Path to the scan results file on the local filesystem scan_data: Raw scan data as string (alternative to file_path) engagement_id: Target engagement ID product_name: Product name (for auto-create context) product_type_name: Product type name (for auto-create context) engagement_name: Engagement name (for auto-create context) test_title: Custom test title auto_create_context: Auto-create product/engagement if they don't exist active: Mark findings as active verified: Mark findings as verified close_old_findings: Close findings not present in new scan close_old_findings_product_scope: Close at product scope push_to_jira: Push findings to JIRA minimum_severity: Minimum severity to import (Info, Low, Medium, High, Critical) deduplication_on_engagement: Deduplicate within engagement environment: Environment name version: Version being scanned build_id: Build ID branch_tag: Branch or tag commit_hash: Commit hash service: Service name group_by: Group findings by (component_name, component_name+component_version, file_path) tags: Tags to apply

Inputs: scan_type (required), tags, active, service, +20 more. Returns: object

list_endpointsB

List endpoints with optional filters.

Args: product_id: Filter by product ID host: Filter by host (contains) protocol: Filter by protocol path: Filter by path (contains) tag: Filter by tag limit: Results per page offset: Pagination offset

Inputs: tag, host, path, limit, +3 more. Returns: object

list_endpoint_statusC

List endpoint statuses (relationship between endpoints and findings).

Args: endpoint_id: Filter by endpoint ID finding_id: Filter by finding ID mitigated: Filter by mitigated status limit: Results per page offset: Pagination offset

Inputs: limit, offset, mitigated, finding_id, +1 more. Returns: object

list_engagement_notesC

List notes for an engagement.

Args: engagement_id: The engagement ID

Inputs: engagement_id (required). Returns: object

list_engagementsB

List engagements with optional filters.

Args: product_id: Filter by product ID name: Filter by name (contains) status: Filter by status (Not Started, In Progress, Completed) engagement_type: Filter by type (Interactive, CI/CD) tag: Filter by tag name limit: Results per page (default 25) offset: Pagination offset

Inputs: tag, name, limit, offset, +3 more. Returns: object

list_finding_metadataC

List metadata key-value pairs for a finding.

Args: finding_id: The finding ID

Inputs: finding_id (required). Returns: object

list_finding_notesB

List notes for a finding (ordered newest first).

Args: finding_id: The finding ID

Inputs: finding_id (required). Returns: object

list_findingsB

List findings with optional filters.

Args: severity: Filter by severity (Critical, High, Medium, Low, Info) active: Filter active findings only verified: Filter verified findings only is_mitigated: Filter mitigated findings duplicate: Filter duplicates false_positive: Filter false positives (false_p) out_of_scope: Filter out-of-scope findings risk_accepted: Filter findings under an active risk acceptance test_id: Filter by test ID engagement_id: Filter by engagement ID (via test__engagement) product_id: Filter by product (asset) ID (via test__engagement__product) product_name: Filter by product (asset) name (server-side contains) title: Filter by title (server-side contains) title_exact: Filter by exact title cwe: Filter by CWE number vulnerability_id: Filter by CVE/vulnerability ID (e.g. CVE-2024-1234) reporter_id: Filter by reporter user ID mitigated_by_id: Filter by the user who mitigated outside_of_sla: Filter findings outside SLA tag: Filter by tag name (contains) ordering: Ordering fields (e.g. "-severity", "created", "-date") limit: Number of results per page (default 25) offset: Pagination offset

Inputs: cwe, tag, limit, title, +19 more. Returns: object

list_jira_finding_mappingsA

List JIRA finding mappings (links between findings and JIRA issues).

Args: finding_id: Filter by finding ID jira_key: Filter by JIRA issue key limit: Results per page offset: Pagination offset

Inputs: limit, offset, jira_key, finding_id. Returns: object

list_jira_instancesC

List configured JIRA instances.

Args: limit: Results per page offset: Pagination offset

Inputs: limit, offset. Returns: object

list_jira_projectsC

List JIRA project configurations.

Args: product_id: Filter by product ID engagement_id: Filter by engagement ID limit: Results per page offset: Pagination offset

Inputs: limit, offset, product_id, engagement_id. Returns: object

list_note_typesA

List available note types (usable in add_finding_note / close_finding).

Args: name: Filter by name (contains) limit: Results per page (default 25) offset: Pagination offset

Inputs: name, limit, offset. Returns: object

list_notificationsC

List notification configurations.

Args: limit: Results per page offset: Pagination offset

Inputs: limit, offset. Returns: object

list_organizationsB

List organizations (product types).

Args: name: Filter by name (contains) limit: Results per page (default 25) offset: Pagination offset

Inputs: name, limit, offset. Returns: object

list_productsB

List products (assets) with optional filters.

Args: name: Filter by name (contains) name_exact: Filter by exact name organization_id: Filter by organization (product type) ID lifecycle: Filter by lifecycle (construction, production, retirement) tag: Filter by tag name (contains) external_audience: Filter by external audience flag internet_accessible: Filter by internet accessible flag limit: Results per page (default 25) offset: Pagination offset

Inputs: tag, name, limit, offset, +5 more. Returns: object

list_risk_acceptancesB

List risk acceptances.

Args: finding_id: Filter by an accepted finding ID owner_id: Filter by owner user ID name_contains: Filter by name (contains) accepted_by: Filter by acceptor name/email (contains) limit: Results per page offset: Pagination offset

Inputs: limit, offset, owner_id, finding_id, +2 more. Returns: object

list_sla_configurationsC

List SLA configurations.

Args: limit: Results per page offset: Pagination offset

Inputs: limit, offset. Returns: object

list_test_notesC

List notes for a test.

Args: test_id: The test ID

Inputs: test_id (required). Returns: object

list_testsB

List tests with optional filters.

Args: engagement_id: Filter by engagement ID product_id: Filter by product ID (via engagement__product) test_type: Filter by test type ID tag: Filter by tag name (contains) title: Filter by title (contains) limit: Results per page (default 25) offset: Pagination offset

Inputs: tag, limit, title, offset, +3 more. Returns: object

list_test_typesC

List available test types (scan types).

Args: name: Filter by name (contains) limit: Results per page offset: Pagination offset

Inputs: name, limit, offset. Returns: object

list_usersB

List users with optional filters.

Args: username: Filter by username (contains) first_name: Filter by first name last_name: Filter by last name is_active: Filter by active status is_superuser: Filter by superuser status limit: Results per page offset: Pagination offset

Inputs: limit, offset, username, is_active, +3 more. Returns: object

purge_celery_queueA

Purge all pending Celery tasks from the queue.

Tasks already being executed are not affected. Use with caution - this removes ALL pending tasks.

Inputs: no named parameters. Returns: object

purge_celery_taskB

Purge all queued instances of a specific Celery task.

Args: task_name: The task name to purge (e.g., "dojo.tasks.async_update")

Inputs: task_name (required). Returns: object

reimport_scanA

Reimport a scan report (update existing test with new results).

Either test_id OR (product_name + test_title + auto_create_context) must be provided. Either file_path or scan_data must be provided.

Args: scan_type: Scanner type (e.g., "ZAP Scan", "Nessus Scan", "SARIF", etc.) test_id: Existing test ID to reimport into file_path: Path to the scan results file scan_data: Raw scan data as string product_name: Product name (for auto-create context) product_type_name: Product type name (for auto-create context) engagement_name: Engagement name (for auto-create context) test_title: Test title (for matching existing test) auto_create_context: Auto-create context if it doesn't exist active: Mark findings as active verified: Mark findings as verified close_old_findings: Close findings not in new scan (default True) close_old_findings_product_scope: Close at product scope push_to_jira: Push to JIRA minimum_severity: Minimum severity do_not_reactivate: Don't reactivate closed findings environment: Environment name version: Version being scanned build_id: Build ID branch_tag: Branch or tag commit_hash: Commit hash service: Service name group_by: Group findings by field tags: Tags to apply

Inputs: scan_type (required), tags, active, service, +20 more. Returns: object

reinstate_risk_acceptanceA

Reinstate an expired risk acceptance with a new (mandatory) expiration date.

Args: risk_acceptance_id: The risk acceptance ID expiration_date: New expiration date (YYYY-MM-DD or ISO datetime), must be in the future reason: Optional reason

Inputs: expiration_date (required), risk_acceptance_id (required), reason. Returns: object

remove_finding_noteC

Remove a note from a finding.

Args: finding_id: The finding ID note_id: The note ID to remove (see list_finding_notes)

Inputs: note_id (required), finding_id (required). Returns: object

reopen_findingA

Reopen (reactivate) a closed/mitigated finding.

Clears the mitigated state and any false-positive/duplicate/out-of-scope flags unless explicitly restored.

Args: finding_id: The finding ID to reopen note: Optional note explaining the reopening restore_false_positive: Keep the false_p flag as-is instead of clearing it restore_out_of_scope: Keep the out_of_scope flag as-is instead of clearing it

Inputs: finding_id (required), note, restore_out_of_scope, restore_false_positive. Returns: object

reset_finding_duplicate_statusC

Reset the duplicate status of a finding.

Args: finding_id: The finding ID

Inputs: finding_id (required). Returns: object

update_engagementB

Update an existing engagement (partial update).

Args: engagement_id: The engagement ID name: New name status: New status (Not Started, In Progress, Completed) description: New description target_start: New target start date target_end: New target end date tags: Replace tags

Inputs: engagement_id (required), name, tags, status, +3 more. Returns: object

update_findingC

Update an existing finding (partial update).

Args: finding_id: The finding ID title: New title severity: New severity description: New description mitigation: New mitigation text impact: New impact text active: Set active status verified: Set verified status is_mitigated: Set mitigated status tags: Replace tags

Inputs: finding_id (required), tags, title, active, +6 more. Returns: object

update_organizationB

Update an organization (partial update).

Args: organization_id: The organization ID name: New name description: New description

Inputs: organization_id (required), name, description. Returns: object

update_productB

Update an existing product (partial update).

Args: product_id: The product ID name: New name description: New description lifecycle: New lifecycle stage business_criticality: New business criticality tags: Replace tags

Inputs: product_id (required), name, tags, lifecycle, +2 more. Returns: object

update_risk_acceptanceB

Update a risk acceptance (partial update, e.g. extend expiration).

Args: risk_acceptance_id: The risk acceptance ID expiration_date: New expiration date (YYYY-MM-DD or ISO datetime) decision: Risk treatment decision code (A, V, M, F, T) decision_details: Details about the decision recommendation: Security recommendation code (A, V, M, F, T) recommendation_details: Details about the recommendation reactivate_expired: Reactivate findings on expiration restart_sla_expired: Restart SLA on expiration

Inputs: risk_acceptance_id (required), decision, recommendation, expiration_date, +4 more. Returns: object

update_system_settingsC

Update system settings (partial update).

Args: settings_id: Settings ID (usually 1) **kwargs: Key-value pairs of settings to update

Inputs: kwargs (required), settings_id. Returns: object

update_testC

Update an existing test (partial update).

Args: test_id: The test ID title: New title description: New description version: New version tags: Replace tags

Inputs: test_id (required), tags, title, version, +1 more. Returns: object

update_userC

Update a user (partial update).

Args: user_id: The user ID first_name: New first name last_name: New last name email: New email is_active: Set active status

Inputs: user_id (required), email, is_active, last_name, +1 more. Returns: object

verify_findingC

Mark a finding as verified.

Args: finding_id: The finding ID to verify

Inputs: finding_id (required). Returns: object

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    C2.9/5.0

    Scored across 78 tools

    Disambiguation3/5

    Most tools are clearly distinct by resource and action (e.g., list_products vs get_product). However, the closure tools are heavily overlapping: close_finding, close_finding_false_positive, close_finding_mitigated, and close_finding_duplicate all serve the same core purpose with different levels of abstraction. Similarly, accept_risk, accept_risks_vulnerability, and create_risk_acceptance overlap, requiring agents to choose carefully.

    Naming Consistency4/5

    The naming is mostly consistent with a verb_noun pattern (list_products, create_product, delete_product). Some deviations exist, such as import_scan and reimport_scan (no noun), and tools like get_current_user follow the pattern. The heavy use of 'finding' variations (add_finding_metadata, close_finding_duplicate) is predictable but occasionally verbose.

    Tool Count2/5

    With 78 tools, the server is far beyond the recommended 3-15 range for a single MCP server. This is an extreme mismatch that makes it difficult for an agent to navigate and increases the risk of misselection. While the domain (DefectDojo) is complex, the count is excessive and could be split into multiple focused servers.

    Completeness4/5

    The tool surface covers the core domain comprehensively: products, engagements, tests, findings, users, risk acceptances, and JIRA integrations. Minor gaps exist, such as no explicit tool to list all available scan types (list_test_types exists but may not cover external scanners), and some bulk operations are missing (e.g., bulk update findings). However, most CRUD operations are present.

    Maintenance

    ActivityStale
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      B
      quality
      D
      maintenance
      Provides a Model Context Protocol server implementation that allows AI agents and other MCP clients to programmatically interact with DefectDojo, a vulnerability management tool, for managing findings, products, and engagements.
      11
      107 PyPI
      21
      MIT
    • A
      license
      A
      quality
      D
      maintenance
      MCP server for DefectDojo vulnerability management, exposing 24 tools for managing products, engagements, tests, findings, scan imports, and finding lifecycle through the Model Context Protocol.
      24
      1
      MIT