How is this different from connecting Dynatrace MCP directly?

Connected directly, an agent sees every tool the server exposes and holds the credential itself. Through TrustGate you choose which tools each agent gets, and every call is logged against the agent that made it.

What if the Dynatrace MCP API changes?

We track the Dynatrace MCP schema, so the tool catalogue stays current. Tools you have already allowed keep working, and anything new stays denied until you enable it.

Can I give different agents different Dynatrace tools?

Yes. Access is set per agent: one can get a handful of read tools, another the full set. You grant the whole MCP or single tools within it.

Where can I see what my agents did in Dynatrace?

Every call is logged against the agent that made it, with the tool it used. Activity can be sent on to your own monitoring.

Do I need to run the Dynatrace MCP server myself?

No. TrustGate connects to the Dynatrace MCP server and sits in front of it. Your agent only talks to your TrustGate URL, never to Dynatrace directly.

Connect Dynatrace MCP to your AI agents, securely.

Q: Is Dynatrace MCP secure?

The server itself exposes 17 tools, and some of them can write to your Dynatrace account. Behind TrustGate an agent only sees the tools you allow, requests and responses are inspected, and every call is logged. Anything outside the allowlist is denied before it reaches Dynatrace.

One secured endpoint for every Dynatrace call your AI agents make, scoped per user, inspected, and logged.

Get started for free

Get a demo

Try Dynatrace MCP now

What your agents can ask once Dynatrace MCP is connected.

Select the tools your AI agents can call

17 tools from the Dynatrace MCP schema.

Showing 17 tools

create-dqlreadGrail Query Agent. Generates a DQL query from a natural-language prompt. It doesn't execute the query. Requires the davis-copilot:nl2dql:execute permission.
explain-dqlreadDQL Explanation Agent. Provides a natural-language explanation of a DQL query. Requires the davis-copilot:dql2nl:execute permission.
execute-dqlhigh riskData Analysis Agent. Executes any valid DQL query and returns the raw result, limited to 1000 Grail records. Requires storage:buckets:read plus the storage permissions for the data queried, such as storage:logs:read, and respects those scopes.
query-problemsreadRoot Cause Agent. Returns an overview list of all problems in the environment, active or closed. Requires storage:buckets:read and storage:events:read.
get-problem-by-idreadRoot Cause Details Agent. Returns details of a specific Davis problem. Requires storage:buckets:read and storage:events:read.
get-events-for-kubernetes-clusterreadKubernetes Agent. Returns events for all Kubernetes clusters or specific ones. Requires storage:buckets:read and storage:events:read.
timeseries-forecastreadForecasting Agent. Predicts future timeseries values using a statistical forecasting model. Requires storage:buckets:read, davis:analyzers:read and davis:analyzers:execute.
timeseries-novelty-detectionreadChangepoint Agent. Analyzes timeseries data to find specific events, outliers and significant trends. Requires storage:buckets:read, davis:analyzers:read and davis:analyzers:execute.
static-threshold-analyzerreadStatic Threshold Analysis Agent. Detects anomalies by setting fixed thresholds for metrics that shouldn't be violated. Requires storage:buckets:read, davis:analyzers:read and davis:analyzers:execute.
seasonal-baseline-anomaly-detectorreadSeasonal Baseline Agent. Detects anomalies in time series by learning normal behavior from historical data and establishing dynamic baseline thresholds that account for seasonal patterns such as daily and weekly traffic.
adaptive-anomaly-detectorreadAuto-adaptive Threshold Analysis Agent. Detects anomalies in time series data by learning a single threshold from the data distribution that separates outliers from normal observations.
ask-dynatrace-docsreadHelp Agent. Answers general Dynatrace questions, for example how to create workflows or what a custom alert is. Requires the davis-copilot:conversations:execute permission.
find-documentsreadDocument Agent. Finds user-accessible Dashboards and Notebooks by name. Requires the document:documents:read permission.
find-troubleshooting-guidesreadTroubleshooting Agent. Finds globally shared troubleshooting guide Dashboards and Notebooks based on a problem description. Needs document suggestion enabled in Settings and the davis-copilot:document-search:execute permission.
get-entity-idreadSmartscape Agent. Looks up the ID of an entity from its name and type. Requires the storage:entities:read permission.
get-entity-namereadSmartscape Agent. Resolves an entity ID to its display name. Requires the storage:entities:read permission.
get-vulnerabilitiesreadList vulnerabilities

What Dynatrace MCP exposes

Dynatrace MCP groups its 17 tools into 6 categories, each with a suggested starting policy.

All 17 tools as JSON, ready to paste.

Category	Covers	Access	Suggested policy
Query	Generating a DQL query from a natural-language prompt, explaining a DQL query in plain language, and executing any valid DQL query against Grail with results capped at 1000 records	Read / Write	create-dql and explain-dql only draft or describe a query. execute-dql runs arbitrary DQL, so what it reaches is set by the storage:*:read permissions on the user and token, not by the tool. Log the query text, keep the token's storage scopes to the data the agent needs, and remember Grail queries consume capacity
Problems	Listing active or closed problems in the environment, reading the details of a specific Davis problem, and retrieving events for all or specific Kubernetes clusters	Read	Read-only incident context and the core of most triage workflows. Safe to allow for on-call and SRE agents whose token carries storage:events:read
Time series analysis	Forecasting timeseries values, finding changepoints and outliers, detecting anomalies with static, seasonal baseline or auto-adaptive thresholds	Read	Read-only Davis analyzers that run over data the agent supplies or queries. They need davis:analyzers:execute on top of storage access
Security	Listing open runtime vulnerabilities	Read	Read-only, but this is a map of what is exploitable in your estate. Keep storage:security.events:read off general-purpose agents and limit this tool to security workflows
Knowledge	Answering product questions about Dynatrace, finding Dashboards and Notebooks the user can access by name, and finding shared troubleshooting guides for a problem description	Read	Low risk and safe to allow broadly. find-documents is limited to documents the user can already open, and find-troubleshooting-guides only works when document suggestion is enabled in Settings
Entities	Looking up an entity ID from its name and type, and resolving an entity ID to its display name through Smartscape	Read	Read-only lookups that other tools depend on to target a host, service or cluster. Safe to allow wherever storage:entities:read is granted

Connect Dynatrace MCP to your AI agent

Four steps, under ten minutes.

01
Sign in to NeuralTrust
Open the NeuralTrust platform, sign in and go to Getting started, then Connect your tools.
02
Choose your agent
When asked which agent you use, pick Claude Code, Codex, Cursor, Windsurf, Copilot or your own code.
03
Connect Dynatrace
Find Dynatrace in the tool list and connect it.
04
Connect your agent
Follow the steps on screen to add the config to your agent, then send the test prompt.

Permissions down to the single tool

You do not allow Dynatrace MCP or block it. You allow the exact tools each agent can call, one by one.

Dynatrace MCP	17 TOOLS AVAILABLE
Code review agent	4 tools, read only
Issue triage bot	9 tools
Platform team assistant	all 17 tools

Tool by toolConnecting the server is not the unit of access. Each tool is granted on its own.
Default denyA tool you have not enabled is never exposed to the agent.
Approval on writesTools that change or delete things can require a human to confirm.

Is Dynatrace MCP secure?

One endpoint instead of raw credentials. Every call is checked against policy and logged.

AI agentsConnect through your MCP client configuration.	NeuralTrustApply controls based on your configuration.	Dynatrace MCP17 tools on Dynatrace's hosted MCP server, served from each SaaS environment at {environment-name}.apps.dynatrace.com/platform-reserved/mcp-gateway/v0.1/servers/dynatrace-mcp/mcp. Every request carries a bearer token from a Platform token or a confidential OAuth client; dynamic client registration is not supported. Tool identifiers as published in the Obot MCP catalog, mapped to the agents in Dynatrace's docs.
Coding assistants	Connection & credentialsper user	Time series analysis5 tools
Internal copilots	Tool access policiesdeny by default	Query3 tools
Workflow automations	Request & response inspectioninline	Problems3 tools
Compatible MCP clients	Activity monitoringalerts and logs	All other tools6 tools