Capabilities and use cases
Last updated: Nov 9, 2026
The Fluid Attacks MCP server has 31 tools. This page shows them by category. You must authenticate to use most tools. Five documentation and scanner tools work without authentication. For the OAuth or API token configuration of each client, read the installation prerequisites.
(a) Start chats with a mention of the API of Fluid Attacks (for example, "Using Fluid Attacks' API, [...]"). (b) Always say if your requests are at the group or the organization level. (c) For requests at the group level, also give the name of the organization.
Documentation
Search articles
Tool: search_related_articles
Searches the documentation of Fluid Attacks for the parts that are related to your request.
When to use:
- To learn about vulnerability types
- To find remediation guidance
- To understand security concepts
Example requests:
- "How to configure the SCA scanner to run on Azure DevOps?"
- "What is cross-site scripting?"
Get articles
Tool: get_articles
Gets the full content of articles from the documentation of Fluid Attacks by article ID (a maximum of 20 at a time).
When to use:
- To read the full articles that
search_related_articlesfinds - To get detailed documentation about a topic
Example requests:
- "Get me the full article on organization policies."
Secure coding context
Get Secure Context rules
Tool: get_secure_context_rules
Loads the security rules that an AI agent must obey when it writes code in your repository. The rules include a set of standard rules and contextual rules from the vulnerability history of your group.
The system finds the Fluid Attacks group of your repository from its remote URL. You do not select the group manually. If the repository is not registered in Fluid Attacks, or if the workspace is not a Git repository, you get only the standard rules. For this, you must have one of these roles on at least one group: Customer Manager, Group Manager, User, or Vulnerability Manager. Without one of these roles, you get a "not authorized" message and no rules.
When to use:
- At the start of a coding session
- Before an AI agent writes or changes code in a repository
- When you want the agent to obey rules from your vulnerability history
Example requests:
- "Load the Fluid Attacks Secure Context rules for this repository."
- "Apply Fluid Attacks security rules before you write this code."
Example output:
[Fluid Attacks Secure Context]
4 rules loaded from Fluid Attacks · fluidattacks.com
(1 from your group's findings · 3 standard)
- Use parameterized queries (CWE-89 · CWE-130) [from your group's
findings] — The system must use parameterized queries or stored
procedures to create dynamic sentences (e.g., java.sql.PreparedStatement).
- Avoid client-side control enforcement (CWE-284 · CWE-285) — The system
must enforce access controls on trusted enforcement points, which are
not on the client's side.
- Encrypt sensitive information (CWE-311 · CWE-526) — All stored
sensitive information must be encrypted.
- Avoid deserializing untrusted data (CWE-95 · CWE-502) — The system must
not deserialize untrusted data before applying the appropriate
integrity checks.The agent shows you this block without changes. Then, it obeys these rules until the session ends.
The contextual rules come from the vulnerability history of your group. Fluid Attacks never mixes rules from different groups in the same response.
For the full guide about Secure Context, with its prerequisites and limits, read Secure Context.
Security scanning
To use the SCA and SAST scanners, you must have Docker. For more information, read "Docker installation".
Run SCA scanner
Tool: run_sca_scanner
Gives instructions to scan the dependencies of your project for known vulnerabilities.
When to use:
- After you add new dependencies
- Before you release new versions
- For regular security audits of libraries
Example requests:
- "How do I scan my dependencies?"
- "Run an SCA scan on this project."
- "Check my npm packages for vulnerabilities."
Run SAST scanner
Tool: run_sast_scanner
Gives instructions to scan your source code for security vulnerabilities.
When to use:
- After you write new code
- Before code reviews
- For security checks before deployment
Example requests:
- "How do I scan my code for vulnerabilities?"
- "Run a SAST scan on the authentication module."
- "Check my Python code for security issues."
Run both scanners
Tool: run_sca_and_sast_scanners
Gives instructions to run SAST and SCA scans at the same time.
When to use:
- For a full security assessment
- As a security gate before production
- For weekly security checks
Example requests:
- "Run a complete security scan."
- "Scan both my code and dependencies."
Run secrets scanner
Tool: run_secrets_scanner
Gives instructions to scan your source code for hardcoded secrets,
API keys, tokens, and credentials.
This scanner is different from the SCA and SAST scanners.
It runs as a self-contained native binary
(the ss binary, which you get with curl).
Docker is not necessary for it.
When to use:
- After you add or change source, configuration, or infrastructure files
- Before you commit or push changes
- To find exposed credentials in a codebase
Example requests:
- "Scan my code for hardcoded secrets."
- "Check this project for exposed API keys or tokens."
Analytics and reporting
Get organization analytics
Tool: get_organization_analytics
Shows security trends and metrics for your full organization.
When to use:
- For monthly security reviews
- For executive reports
- To monitor the security improvement over time
Example requests:
- "Show me our ORGANIZATION_NAME's security risk trend for the last quarter."
- "How has our vulnerability count changed over time in the group GROUP_NAME?"
Available document types:
| Document name | Description |
|---|---|
riskOverTimeCvssf | Risk trend using CVSSF scoring |
riskOverTime | Risk trend using raw counts |
distributionOverTimeCvssf | Vulnerability distribution over time (CVSSF) |
distributionOverTime | Vulnerability distribution over time |
assignedVulnerabilitiesCvssf | Assigned vulnerabilities by CVSSF |
assignedVulnerabilities | Assigned vulnerabilities count |
acceptedVulnerabilitiesBySeverity | Accepted vulnerabilities grouped by severity |
acceptedVulnerabilitiesByUser | Accepted vulnerabilities grouped by user |
cvssfBenchmarking | CVSSF benchmarking across groups |
remediatedAcceptedGroup | Remediated vs accepted by group |
Get group analytics
Tool: get_group_analytics
Shows security trends and metrics for a group (project).
When to use:
- For the security reviews of a project
- For sprint planning
- To monitor how much remediation your team does
Example requests:
- "Show the security trend for the group GROUP_NAME."
- "How many vulnerabilities were fixed in the group GROUP_NAME this month?"
- "What is the distribution of vulnerabilities in the group GROUP_NAME?"
Available document types: The same as the organization analytics, but not cvssfBenchmarking and remediatedAcceptedGroup.
Vulnerability management
Find weaknesses in a group
Tool: fetch_group_weaknesses
Lists the types of security problems (for example, SQL injection and XSS) in your projects. You can filter them by title, status, severity, technique, and location.
When to use:
- To know the types of vulnerabilities that you have
- To plan security training on those types
- To prioritize the remediation work
Example requests:
- "What types of vulnerabilities exist in the group GROUP_NAME?"
- "Show me all critical security weaknesses in the group GROUP_NAME."
Find specific vulnerabilities
Tool: fetch_weakness_vulnerabilities
Lists the vulnerabilities of a weakness, with their locations. You can sort them and filter them by state, treatment, severity, technique, tags, and location.
When to use:
- To assign fixes to developers
- To monitor security problems
- To know where the problems are in your code
Example requests:
- "Show me all XSS vulnerabilities in the frontend code repo for the group GROUP_NAME."
- "List untreated vulnerabilities by priority for the group GROUP_NAME."
Get vulnerability details
Tool: get_vulnerability_details
Shows all the information about a vulnerability, for example, its severity, location, and treatment.
When to use:
- To examine a security problem
- To know how to fix a vulnerability
- To examine a vulnerability before you close it
Example requests:
- "Show me details for vulnerability VULNERABILITY_UUID."
- "What is the full information on the SQL injection in login.py for the group GROUP_NAME?"
Get fix metadata
Tool: get_fix_metadata
Gets remediation guidance for a vulnerability, for the technology of your code.
When to use:
- To get fix instructions for a vulnerability
- To know what code changes are necessary
Example requests:
- "How do I fix vulnerability VULNERABILITY_UUID?"
Get a custom fix
Tool: get_custom_fix
Gets a remediation from AI for a code vulnerability.
The tool generates it from the vulnerable code when you request it.
This can be slow.
It works only for code vulnerabilities.
First, examine the isAutoFixable value.
For SCA (dependency) fixes, use get_fix_metadata.
When to use:
- To get a fix for a code vulnerability that you can apply immediately
Example requests:
- "Get the custom fix for vulnerability VULNERABILITY_UUID."
Get a vulnerability report
Tool: get_group_weaknesses_report
Creates a summary report of the weaknesses, by vulnerability type. You can filter them by title, state, severity, technique, and location.
When to use:
- For security audits
- For compliance reports
- For sprint planning meetings
- For management presentations
Example requests:
- "Generate a vulnerability report for the ProductionAPI group."
- "Show me a report of all open security issues in the group GROUP_NAME."
- "Create a summary of vulnerabilities by severity in the group GROUP_NAME."
Get a vulnerability overview
Tool: get_group_weaknesses_overview
Shows a short summary of the number of vulnerabilities and weaknesses.
When to use:
- For short daily status checks
- To create dashboards
- To set priorities
Example requests:
- "How many open vulnerabilities do we have in the group GROUP_NAME?"
- "Give me a vulnerability count summary for the group GROUP_NAME."
Get vulnerabilities by root
Tool: fetch_group_root_vulnerabilities
Shows the vulnerabilities in a repository or URL. You can filter them by state, severity, and technique.
When to use:
- For the security review of a repository
- To assign work to the teams that are responsible for some assets
- To know the security problems in one part of your infrastructure
Example requests:
- "What vulnerabilities are in the main-app repository?"
- "Show me security issues for the production URL."
Asset discovery
View Git repositories
Tool: get_group_git_roots
Lists the source code repositories that Fluid Attacks tests. You can filter them by nickname and state (active or inactive), or search them.
When to use:
- To see which code Fluid Attacks scans
- To find repositories
- To make sure that the tests include all your code
Example requests:
- "What Git repositories are we testing?"
- "Show me all code repositories for the group GROUP_NAME."
View URLs
Tool: get_group_url_roots
Lists the live websites and web applications that Fluid Attacks tests.
When to use:
- To examine the testing scope in production
- For the security assessment of web applications
Example requests:
- "What websites are we testing?"
- "Show me all URLs for the E-commerce project."
View environment URLs
Tool: get_group_environment_urls
Lists the deployed environment URLs (web, API, CSPM, mobile) for each Git repository in a group, with the nickname, ID, and state of each root.
When to use:
- To find the repository of a deployed environment URL
- To see which environments are registered for a group
Example requests:
- "What environment URLs are registered for the group GROUP_NAME?"
- "Which repository owns the URL https://app.example.com?"
Organization and group management
List organization groups
Tool: get_organization_groups
Lists all the groups of an organization.
Example requests:
- "What groups are in the ORGANIZATION_NAME organization?"
Get organization groups information
Tool: get_organization_groups_information
Gets detailed information about the groups, with the number of weaknesses and vulnerabilities.
Example requests:
- "Show me a summary of all groups in ORGANIZATION_NAME."
Get group information
Tool: get_group_information
Gets the details of a group, for example, its description, language, service level, and managed status.
Example requests:
- "What is the configuration of the group GROUP_NAME?"
Get group remediation information
Tool: get_group_remediation_information
Gets remediation metrics, for example, the mean time to remediate and the remediation rate.
Example requests:
- "What is the remediation rate for the group GROUP_NAME?"
DevSecOps integration
View CI/CD security results
Tool: get_devsecops_agent_executions
Shows the results of the automated security scans in your development pipeline.
When to use:
- To see the build status
- To examine failed security gates
- To monitor if the automation gives good results
Example requests:
- "How many builds failed security checks this week for the group GROUP_NAME?"
- "What was the result of the last Forces execution for the group GROUP_NAME?"
View unsolved events
Tool: get_unsolved_events
Lists the security incidents and conditions that you must examine.
When to use:
- For the daily review of incidents
- For security event management
- To prioritize the problems that you must resolve first
Example requests:
- "What security events need my attention?"
- "Show me unsolved incidents."
Platform navigation
Execute GraphQL queries
Tool: query
Runs any read-only GraphQL query on the Fluid Attacks API. It does not support mutations.
When to use:
- To get data that the other tools do not get
- For advanced or custom queries
Describe GraphQL types
Tool: describe_graphql_type
Examines the GraphQL schema to find the available types, fields, and arguments.
When to use:
- To examine the API schema before you write queries
- To know the available fields of a type
Get user info
Tool: get_user_info
Gets information about the authenticated user, for example, the role, organizations, and groups.
Identify repository context
Tool: identify_repository_context
Finds the Fluid Attacks group and root of a repository URL or nickname.
When to use:
- To find the group of a repository
- To set the context for the next queries
Specialized prompts
The MCP server has specialized prompts that help AI agents do security tasks.
run_sca
Gives the LLM the instructions to run the SCA scanner.
Usage: /fluidattacks-mcp/run_sca
run_sast
Gives the LLM the instructions to run the SAST scanner.
Usage: /fluidattacks-mcp/run_sast
run_sca_and_sast
Gives the LLM the instructions to run the SCA and SAST scanners.
Usage: /fluidattacks-mcp/run_sca_and_sast
run_secrets
Gives the LLM the instructions to run the secrets scanner.
Usage: /fluidattacks-mcp/run_secrets
get_technology_based_remediation_strategy
Gives the LLM the remediation guidance for your technology.
Usage: /fluidattacks-mcp/get_technology_based_remediation_strategy
configure_github_sca_integration
Gives the LLM the instructions to configure the SCA scanner on GitHub Actions.
Usage: /fluidattacks-mcp/configure_github_sca_integration
configure_github_sast_integration
Gives the LLM the instructions to configure the SAST scanner on GitHub Actions.
Usage: /fluidattacks-mcp/configure_github_sast_integration
config_agents_md
Configures the AGENTS.md file in the root directory of the project.
Usage: /fluidattacks-mcp/config_agents_md
How to use these prompts
To use the MCP prompts of Fluid Attacks,
type the command name in the chat of your AI agent.
For example,
to configure the AGENTS.md file,
type /fluidattacks-mcp/config_agents_md.

Select the command that you want to use. Then, press Enter.
Do you have an idea to simplify the architecture of Fluid Attacks? Did you find documentation that Fluid Attacks can make better? Open an issue or send your improvements.
Installation
Integrate the Fluid Attacks MCP server to use natural language queries in AI tools (examples in Claude, Cursor, VS Code, and Windsurf). Also, integrate our scanners into your SDLC using AI agents.
Peer Reviewer Assistant
What the Peer Reviewer Assistant does in a pull request: the review of the lines that the change adds, the summary comment, the status check, the severity threshold, the exclusions, the exceptions, and the record in the platform.