Common Weakness Enumeration

CWE-1336

Allowed

Improper Neutralization of Special Elements Used in a Template Engine

Abstraction: Base · Status: Incomplete

The product uses a template engine to insert or process externally-influenced input, but it does not neutralize or incorrectly neutralizes special elements or syntax that can be interpreted as template expressions or other code directives when processed by the engine.

346 vulnerabilities reference this CWE, most recent first.

GHSA-X7GR-MMJJ-HX3H

Vulnerability from github – Published: 2024-11-14 18:30 – Updated: 2026-04-01 18:32
VLAI
Details

Improper Neutralization of Special Elements Used in a Template Engine vulnerability in Podlove Podlove Podcast Publisher.This issue affects Podlove Podcast Publisher: from n/a through 4.1.15.

Show details on source website

{
  "affected": [],
  "aliases": [
    "CVE-2024-52393"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-1336",
      "CWE-82",
      "CWE-94"
    ],
    "github_reviewed": false,
    "github_reviewed_at": null,
    "nvd_published_at": "2024-11-14T18:15:26Z",
    "severity": "CRITICAL"
  },
  "details": "Improper Neutralization of Special Elements Used in a Template Engine vulnerability in Podlove Podlove Podcast Publisher.This issue affects Podlove Podcast Publisher: from n/a through 4.1.15.",
  "id": "GHSA-x7gr-mmjj-hx3h",
  "modified": "2026-04-01T18:32:24Z",
  "published": "2024-11-14T18:30:38Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2024-52393"
    },
    {
      "type": "WEB",
      "url": "https://patchstack.com/database/Wordpress/Plugin/podlove-podcasting-plugin-for-wordpress/vulnerability/wordpress-podlove-podcast-publisher-plugin-4-1-15-admin-remote-code-execution-rce-vulnerability?_s_id=cve"
    },
    {
      "type": "WEB",
      "url": "https://patchstack.com/database/vulnerability/podlove-podcasting-plugin-for-wordpress/wordpress-podlove-podcast-publisher-plugin-4-1-15-admin-remote-code-execution-rce-vulnerability?_s_id=cve"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:H/UI:N/S:C/C:H/I:H/A:H",
      "type": "CVSS_V3"
    }
  ]
}

GHSA-X7M9-MWC2-G6W2

Vulnerability from github – Published: 2026-05-18 17:23 – Updated: 2026-06-09 10:32
VLAI
Summary
Formie: Pre-authenticated server-side template injection in Hidden fields
Details

Impact

  • Unauthenticated users could submit crafted values into Hidden fields (with Default value → Custom) that were evaluated as Twig during submission handling, which could lead to serious compromise of the Craft site (depending on template/sandbox behavior).
  • Sites with public Formie forms that include at least one Hidden field with that configuration.
  • No CP login for the reported chain.

Patches

Workarounds

  • Temporarily remove Hidden fields from public forms or switch Hidden default away from Custom where feasible
  • Otherwise, upgrade to patched versions
Show details on source website

{
  "affected": [
    {
      "package": {
        "ecosystem": "Packagist",
        "name": "verbb/formie"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "3.0.0-beta.1"
            },
            {
              "fixed": "3.1.24"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "package": {
        "ecosystem": "Packagist",
        "name": "verbb/formie"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "2.2.20"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-45697"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-1336",
      "CWE-693",
      "CWE-94"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-05-18T17:23:39Z",
    "nvd_published_at": "2026-05-29T20:16:27Z",
    "severity": "CRITICAL"
  },
  "details": "### Impact\n- Unauthenticated users could submit crafted values into Hidden fields (with Default value \u2192 Custom) that were evaluated as Twig during submission handling, which could lead to serious compromise of the Craft site (depending on template/sandbox behavior).\n- Sites with public Formie forms that include at least one Hidden field with that configuration.\n- No CP login for the reported chain.\n\n### Patches\n- [2.2.20](https://github.com/verbb/formie/releases/tag/2.2.20), [3.1.24](https://github.com/verbb/formie/releases/tag/3.1.24)\n\n### Workarounds\n- Temporarily remove Hidden fields from public forms or switch Hidden default away from Custom where feasible\n- Otherwise, upgrade to patched versions",
  "id": "GHSA-x7m9-mwc2-g6w2",
  "modified": "2026-06-09T10:32:38Z",
  "published": "2026-05-18T17:23:39Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/verbb/formie/security/advisories/GHSA-x7m9-mwc2-g6w2"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-45697"
    },
    {
      "type": "WEB",
      "url": "https://github.com/verbb/formie/commit/f690d5623163ce2a95da305238d6367575486ee3"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/verbb/formie"
    },
    {
      "type": "WEB",
      "url": "https://github.com/verbb/formie/releases/tag/2.2.20"
    },
    {
      "type": "WEB",
      "url": "https://github.com/verbb/formie/releases/tag/3.1.24"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H",
      "type": "CVSS_V3"
    }
  ],
  "summary": "Formie: Pre-authenticated server-side template injection in Hidden fields"
}

GHSA-XJW8-8C5C-9R79

Vulnerability from github – Published: 2026-04-15 19:46 – Updated: 2026-04-24 20:53
VLAI
Summary
Improper neutralization of specific syntax patterns for unauthorized expressions in Thymeleaf
Details

Impact

A security bypass vulnerability exists in the expression execution mechanisms of Thymeleaf up to and including 3.1.3.RELEASE. Although the library provides mechanisms to prevent expression injection, it fails to properly neutralize specific syntax patterns that allow for the execution of unauthorized expressions. If an application developer passes unvalidated user input directly to the template engine, an unauthenticated remote attacker can bypass the library's protections to achieve Server-Side Template Injection (SSTI).

Patches

This has been fixed in Thymeleaf 3.1.4.RELEASE.

Workarounds

No workaround is available beyond ensuring applications do not pass unvalidated user input directly to the template engine. Upgrading to 3.1.4.RELEASE is strongly recommended in any case.

Credits

Thanks to Dawid Bakaj (VIPentest.com) for responsible disclosure.

Show details on source website

{
  "affected": [
    {
      "database_specific": {
        "last_known_affected_version_range": "\u003c= 3.1.3.RELEASE"
      },
      "package": {
        "ecosystem": "Maven",
        "name": "org.thymeleaf:thymeleaf"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "3.1.4.RELEASE"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "database_specific": {
        "last_known_affected_version_range": "\u003c= 3.1.3.RELEASE"
      },
      "package": {
        "ecosystem": "Maven",
        "name": "org.thymeleaf:thymeleaf-spring5"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "3.1.4.RELEASE"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    },
    {
      "database_specific": {
        "last_known_affected_version_range": "\u003c= 3.1.3.RELEASE"
      },
      "package": {
        "ecosystem": "Maven",
        "name": "org.thymeleaf:thymeleaf-spring6"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "3.1.4.RELEASE"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-40478"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-1336",
      "CWE-917"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-04-15T19:46:23Z",
    "nvd_published_at": "2026-04-17T22:16:33Z",
    "severity": "CRITICAL"
  },
  "details": "### Impact\nA security bypass vulnerability exists in the expression execution mechanisms of Thymeleaf up to and including 3.1.3.RELEASE. Although the library provides mechanisms to prevent expression injection, it fails to properly neutralize specific syntax patterns that allow for the execution of unauthorized expressions. If an application developer passes unvalidated user input directly to the template engine, an unauthenticated remote attacker can bypass the library\u0027s protections to achieve Server-Side Template Injection (SSTI).\n\n### Patches\nThis has been fixed in Thymeleaf 3.1.4.RELEASE.\n\n### Workarounds\nNo workaround is available beyond ensuring applications do not pass unvalidated user input directly to the template engine. Upgrading to 3.1.4.RELEASE is strongly recommended in any case.\n\n### Credits\nThanks to Dawid Bakaj (VIPentest.com) for responsible disclosure.",
  "id": "GHSA-xjw8-8c5c-9r79",
  "modified": "2026-04-24T20:53:26Z",
  "published": "2026-04-15T19:46:23Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/thymeleaf/thymeleaf/security/advisories/GHSA-xjw8-8c5c-9r79"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-40478"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/thymeleaf/thymeleaf"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:C/C:H/I:H/A:H",
      "type": "CVSS_V3"
    }
  ],
  "summary": "Improper neutralization of specific syntax patterns for unauthorized expressions in Thymeleaf"
}

GHSA-XQMJ-J6MV-4862

Vulnerability from github – Published: 2026-04-24 16:02 – Updated: 2026-05-12 13:27
VLAI
Summary
LiteLLM: Server-Side Template Injection in /prompts/test endpoint
Details

Impact

The POST /prompts/test endpoint accepted user-supplied prompt templates and rendered them without sandboxing. A crafted template could run arbitrary code inside the LiteLLM Proxy process.

The endpoint only checks that the caller presents a valid proxy API key, so any authenticated user could reach it. Depending on how the proxy is deployed, this could expose secrets in the process environment (such as provider API keys or database credentials) and allow commands to be run on the host.

Proxy deployments running an affected version are in scope.

Patches

The issue is fixed in 1.83.7-stable. The fix switches the prompt template renderer to a sandboxed environment that blocks the attributes this attack relies on.

LiteLLM recommends upgrading to 1.83.7-stable or later.

Workarounds

If upgrading is not immediately possible:

  1. Block POST /prompts/test at your reverse proxy or API gateway.
  2. Review and rotate API keys that should not have access to prompt management routes.
Show details on source website

{
  "affected": [
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "litellm"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "1.80.5"
            },
            {
              "fixed": "1.83.7"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-42203"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-1336"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-04-24T16:02:42Z",
    "nvd_published_at": "2026-05-08T04:16:19Z",
    "severity": "HIGH"
  },
  "details": "### Impact\nThe `POST /prompts/test` endpoint accepted user-supplied prompt templates and rendered them without sandboxing. A crafted template could run arbitrary code inside the LiteLLM Proxy process.\n\nThe endpoint only checks that the caller presents a valid proxy API key, so any authenticated user could reach it. Depending on how the proxy is deployed, this could expose secrets in the process environment (such as provider API keys or database credentials) and allow commands to be run on the host.\n\nProxy deployments running an affected version are in scope.\n\n### Patches\nThe issue is fixed in **`1.83.7-stable`**. The fix switches the prompt template renderer to a sandboxed environment that blocks the attributes this attack relies on.\n\nLiteLLM recommends upgrading to `1.83.7-stable` or later.\n\n### Workarounds\nIf upgrading is not immediately possible:\n\n1. Block `POST /prompts/test` at your reverse proxy or API gateway.\n2. Review and rotate API keys that should not have access to prompt management routes.",
  "id": "GHSA-xqmj-j6mv-4862",
  "modified": "2026-05-12T13:27:01Z",
  "published": "2026-04-24T16:02:42Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/BerriAI/litellm/security/advisories/GHSA-xqmj-j6mv-4862"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-42203"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/BerriAI/litellm"
    },
    {
      "type": "WEB",
      "url": "https://github.com/BerriAI/litellm/releases/tag/v1.83.7-stable"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:H/VI:H/VA:N/SC:N/SI:N/SA:N",
      "type": "CVSS_V4"
    }
  ],
  "summary": "LiteLLM: Server-Side Template Injection in /prompts/test endpoint"
}

GHSA-XRH7-2GFQ-4RCQ

Vulnerability from github – Published: 2024-07-17 21:31 – Updated: 2024-12-18 22:03
VLAI
Summary
openCart Server-Side Template Injection (SSTI) vulnerability
Details

A Server-Side Template Injection (SSTI) vulnerability in the Theme Editor Function of openCart project v4.0.2.3 allows attackers to execute arbitrary code via injecting a crafted payload.

Show details on source website

{
  "affected": [
    {
      "package": {
        "ecosystem": "Packagist",
        "name": "opencart/opencart"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "last_affected": "4.0.2.3"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2024-36694"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-1336",
      "CWE-94"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2024-07-17T23:36:49Z",
    "nvd_published_at": "2024-07-17T19:15:11Z",
    "severity": "MODERATE"
  },
  "details": "A Server-Side Template Injection (SSTI) vulnerability in the Theme Editor Function of openCart project v4.0.2.3 allows attackers to execute arbitrary code via injecting a crafted payload.",
  "id": "GHSA-xrh7-2gfq-4rcq",
  "modified": "2024-12-18T22:03:16Z",
  "published": "2024-07-17T21:31:38Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2024-36694"
    },
    {
      "type": "WEB",
      "url": "https://github.com/opencart/opencart/issues/13863"
    },
    {
      "type": "WEB",
      "url": "https://github.com/A3h1nt/CVEs/blob/main/OpenCart/Readme.md"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/opencart/opencart"
    },
    {
      "type": "WEB",
      "url": "https://github.com/opencart/opencart/releases/tag/4.0.2.3"
    },
    {
      "type": "WEB",
      "url": "https://medium.com/@pawarit.sanguanpang/opencart-v4-0-2-3-server-side-template-injection-0b173a3bdcf9"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:H/PR:H/UI:N/S:C/C:H/I:H/A:H",
      "type": "CVSS_V3"
    },
    {
      "score": "CVSS:4.0/AV:N/AC:H/AT:N/PR:H/UI:N/VC:N/VI:N/VA:N/SC:H/SI:H/SA:H",
      "type": "CVSS_V4"
    }
  ],
  "summary": "openCart Server-Side Template Injection (SSTI) vulnerability"
}

GHSA-XRMJ-5G4G-8987

Vulnerability from github – Published: 2026-07-31 16:01 – Updated: 2026-07-31 16:01
VLAI
Summary
@dynatrace-oss/dynatrace-mcp-server has a workflow template injection via create_workflow_for_notification
Details

Summary

A template injection vulnerability in the create_workflow_for_notification tool lets a caller embed Jinja2 expressions that the Dynatrace workflow engine evaluates at runtime, exfiltrating event data to attacker-controlled destinations through a workflow that persists in the tenant after the MCP session ends.

Details

The create_workflow_for_notification tool interpolates three caller-supplied parameters (teamName, problemType, channel) directly into a Dynatrace Workflow definition. Dynatrace Workflows use Jinja2 templating: per the official documentation, {{ ... }} expressions in action inputs are evaluated at workflow runtime for every action except Run Javascript (which is carved out specifically to avoid code injection). A caller can therefore supply, for example, teamName = "{{ event() }}" and have the workflow engine evaluate that expression at runtime, serialising the full event object into the message body delivered to the Slack channel.

The vulnerable code is in src/capabilities/create-workflow-for-problem-notification.ts, lines 82-99:

let notificationWorkflow: WorkflowCreate = {
  title: `[MCP POC] Notify team ${teamName} on problem of type ${problemType}`,
  description: `Automatically created workflow to notify team ${teamName} on problems of type ${problemType} - ...`,
  isPrivate: isPrivate,
  type: 'SIMPLE',
  tasks: {
    send_notification: {
      name: 'Send notification',
      action: 'dynatrace.slack:slack-send-message',
      description: 'Sends a notification to a Slack channel',
      input: {
        connectionId: 'slack-connection-id',
        channel: `{{ \"${channel}\" }}`,        // <-- channel sits inside {{ }}
        message: `🚨 Alert for Team ${teamName}\n*Problem Type*: ${problemType}\n` +
                 `*Problem ID*: {{ event()["display_id"] }}\n*Status*: {{ event()["event.status"] }}\n` +
                 `<{{ environment().url }}/ui/apps/.../problem/{{ event()["event.id"] }}|Click here>`,
      },
      active: true,
    },
  },
};

The action used is dynatrace.slack:slack-send-message, which is not in the documented Jinja-expression exception list. Its inputs are evaluated at workflow runtime.

The schema in src/index.ts:1052-1070 registers teamName, problemType, and channel as z.string().optional() with no pattern validation. The isPrivate parameter has .default(false), so created workflows are visible tenant-wide unless the caller explicitly sets it.

The approval prompt at src/index.ts:1069-1072:

const approved = await requestHumanApproval(
  `Create a workflow for notifying team ${teamName} via ${channel} about ${problemType} problems`,
);

Renders {{ event() }} literally to the operator with no indication that it will be templated, and does not surface the workflow's visibility.

The workflow is persistent: it remains in the tenant after the MCP session ends, after the operator's MCP credentials are revoked, and after the MCP server is uninstalled. It fires on every matching problem until manually deleted from the Workflows app. The channel parameter is uniquely dangerous because it is interpolated inside an existing {{ "..." }} expression context - close the string with " and you can run arbitrary Jinja expressions in the destination field itself.

PoC

Tested end-to-end against a real Dynatrace tenant. The MCP server was run in stdio mode with the operator's Platform Token. A tools/call create_workflow_for_notification was sent with:

{
  "teamName":    "{{ event() }}",
  "problemType": "ERROR",
  "channel":     "#mcp-sec-poc",
  "isPrivate":   true
}

The operator approved the prompt (which read: "Create a workflow for notifying team {{ event() }} via #mcp-sec-poc about ERROR problems"). The MCP returned a workflow ID. Fetching the stored workflow body via the Dynatrace Automation API (GET /platform/automation/v1/workflows/<id>) showed the injected expression stored verbatim:

title:   "[MCP POC] Notify team {{ event() }} on problem of type ERROR"
message: "🚨 Alert for Team {{ event() }}\n*Problem Type*: ERROR\n*Problem ID*: {{ event()[\"display_id\"] }}\n..."
channel: '{{ "#mcp-sec-poc" }}'
action:  "dynatrace.slack:slack-send-message"

The workflow was then triggered manually via the "Run workflow" feature in the Dynatrace Workflows app, with a synthetic event payload {"display_id":"P-123","event.status":"OPEN","event.id":"abc-123"}. The execution log for the send_notification task shows the workflow engine evaluated the injected expressions at runtime. The "Input" tab for the executed action contains:

channel: #mcp-sec-poc
message: Alert for Team
         {'display_id': 'P-123',
          'event.status': 'OPEN',
          'event.id': 'abc-123'}
         *Problem Type*: ERROR
         *Problem ID*: P-123
         *Status*: OPEN
         <https://<tenant>.apps.dynatrace.com/ui/apps/.../problem/abc-123|Click here for details>

The injected {{ event() }} resolved to the actual event object before the Slack action was called. The Slack action then failed only because the workflow uses a hardcoded placeholder connectionId: 'slack-connection-id' that does not resolve to any real connection. If a real Slack connector had been configured, the message would have been delivered with the serialized event data in place of {{ event() }}.

This is end-to-end confirmation of the Jinja-evaluation chain.

Impact

A workflow created with a malicious template payload acts as a persistent exfiltration channel:

  • It fires on every matching problem indefinitely.
  • The injected template is evaluated at runtime and resolves to whatever the Jinja function returns (event() gives the full event object, environment() gives tenant metadata, plus other available functions per the Dynatrace Workflows documentation).
  • The resolved output is delivered to a destination the attacker controls (the channel parameter is also templated, and is even more dangerous because it is interpolated inside an existing {{ "..." }} expression context).
  • The workflow persists in the tenant after the MCP session, operator credentials, and MCP server are gone.
  • With isPrivate=false (the default), the workflow is visible to all tenant users.

Exploitation requires either prompt injection (an LLM under the MCP server's control is reading attacker-controlled data and follows an injection that calls create_workflow_for_notification with the malicious arguments) or operator carelessness (the approval prompt shows the raw team name literal and the operator clicks Approve without recognising {{ event() }} as a template fragment). Both are realistic in practice.

Show details on source website

{
  "affected": [
    {
      "package": {
        "ecosystem": "npm",
        "name": "@dynatrace-oss/dynatrace-mcp-server"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "2.0.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [],
  "database_specific": {
    "cwe_ids": [
      "CWE-1336"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-07-31T16:01:07Z",
    "nvd_published_at": null,
    "severity": "MODERATE"
  },
  "details": "### Summary\nA template injection vulnerability in the `create_workflow_for_notification` tool lets a caller embed Jinja2 expressions that the Dynatrace workflow engine evaluates at runtime, exfiltrating event data to attacker-controlled destinations through a workflow that persists in the tenant after the MCP session ends.\n\n### Details\nThe `create_workflow_for_notification` tool interpolates three caller-supplied parameters (`teamName`, `problemType`, `channel`) directly into a Dynatrace Workflow definition. Dynatrace Workflows use Jinja2 templating: per the [official documentation](https://docs.dynatrace.com/docs/analyze-explore-automate/workflows/reference), `{{ ... }}` expressions in action inputs are evaluated at workflow runtime for every action except `Run Javascript` (which is carved out specifically to avoid code injection). A caller can therefore supply, for example, `teamName = \"{{ event() }}\"` and have the workflow engine evaluate that expression at runtime, serialising the full event object into the message body delivered to the Slack channel.\n\nThe vulnerable code is in `src/capabilities/create-workflow-for-problem-notification.ts`, lines 82-99:\n\n```typescript\nlet notificationWorkflow: WorkflowCreate = {\n  title: `[MCP POC] Notify team ${teamName} on problem of type ${problemType}`,\n  description: `Automatically created workflow to notify team ${teamName} on problems of type ${problemType} - ...`,\n  isPrivate: isPrivate,\n  type: \u0027SIMPLE\u0027,\n  tasks: {\n    send_notification: {\n      name: \u0027Send notification\u0027,\n      action: \u0027dynatrace.slack:slack-send-message\u0027,\n      description: \u0027Sends a notification to a Slack channel\u0027,\n      input: {\n        connectionId: \u0027slack-connection-id\u0027,\n        channel: `{{ \\\"${channel}\\\" }}`,        // \u003c-- channel sits inside {{ }}\n        message: `\ud83d\udea8 Alert for Team ${teamName}\\n*Problem Type*: ${problemType}\\n` +\n                 `*Problem ID*: {{ event()[\"display_id\"] }}\\n*Status*: {{ event()[\"event.status\"] }}\\n` +\n                 `\u003c{{ environment().url }}/ui/apps/.../problem/{{ event()[\"event.id\"] }}|Click here\u003e`,\n      },\n      active: true,\n    },\n  },\n};\n```\n\nThe action used is `dynatrace.slack:slack-send-message`, which is not in the documented Jinja-expression exception list. Its inputs are evaluated at workflow runtime.\n\nThe schema in `src/index.ts:1052-1070` registers `teamName`, `problemType`, and `channel` as `z.string().optional()` with no pattern validation. The `isPrivate` parameter has `.default(false)`, so created workflows are visible tenant-wide unless the caller explicitly sets it.\n\nThe approval prompt at `src/index.ts:1069-1072`:\n\n```typescript\nconst approved = await requestHumanApproval(\n  `Create a workflow for notifying team ${teamName} via ${channel} about ${problemType} problems`,\n);\n```\n\nRenders `{{ event() }}` literally to the operator with no indication that it will be templated, and does not surface the workflow\u0027s visibility.\n\nThe workflow is persistent: it remains in the tenant after the MCP session ends, after the operator\u0027s MCP credentials are revoked, and after the MCP server is uninstalled. It fires on every matching problem until manually deleted from the Workflows app. The `channel` parameter is uniquely dangerous because it is interpolated inside an existing `{{ \"...\" }}` expression context - close the string with `\"` and you can run arbitrary Jinja expressions in the destination field itself.\n\n\n### PoC\n\nTested end-to-end against a real Dynatrace tenant. The MCP server was run in stdio mode with the operator\u0027s Platform Token. A `tools/call create_workflow_for_notification` was sent with:\n\n```json\n{\n  \"teamName\":    \"{{ event() }}\",\n  \"problemType\": \"ERROR\",\n  \"channel\":     \"#mcp-sec-poc\",\n  \"isPrivate\":   true\n}\n```\n\nThe operator approved the prompt (which read: `\"Create a workflow for notifying team {{ event() }} via #mcp-sec-poc about ERROR problems\"`). The MCP returned a workflow ID. Fetching the stored workflow body via the Dynatrace Automation API (`GET /platform/automation/v1/workflows/\u003cid\u003e`) showed the injected expression stored verbatim:\n\n```\ntitle:   \"[MCP POC] Notify team {{ event() }} on problem of type ERROR\"\nmessage: \"\ud83d\udea8 Alert for Team {{ event() }}\\n*Problem Type*: ERROR\\n*Problem ID*: {{ event()[\\\"display_id\\\"] }}\\n...\"\nchannel: \u0027{{ \"#mcp-sec-poc\" }}\u0027\naction:  \"dynatrace.slack:slack-send-message\"\n```\n\nThe workflow was then triggered manually via the \"Run workflow\" feature in the Dynatrace Workflows app, with a synthetic event payload `{\"display_id\":\"P-123\",\"event.status\":\"OPEN\",\"event.id\":\"abc-123\"}`. The execution log for the `send_notification` task shows the workflow engine evaluated the injected expressions at runtime. The \"Input\" tab for the executed action contains:\n\n```\nchannel: #mcp-sec-poc\nmessage: Alert for Team\n         {\u0027display_id\u0027: \u0027P-123\u0027,\n          \u0027event.status\u0027: \u0027OPEN\u0027,\n          \u0027event.id\u0027: \u0027abc-123\u0027}\n         *Problem Type*: ERROR\n         *Problem ID*: P-123\n         *Status*: OPEN\n         \u003chttps://\u003ctenant\u003e.apps.dynatrace.com/ui/apps/.../problem/abc-123|Click here for details\u003e\n```\n\nThe injected `{{ event() }}` resolved to the actual event object before the Slack action was called. The Slack action then failed only because the workflow uses a hardcoded placeholder `connectionId: \u0027slack-connection-id\u0027` that does not resolve to any real connection. If a real Slack connector had been configured, the message would have been delivered with the serialized event data in place of `{{ event() }}`.\n\nThis is end-to-end confirmation of the Jinja-evaluation chain.\n\n### Impact\nA workflow created with a malicious template payload acts as a persistent exfiltration channel:\n\n- It fires on every matching problem indefinitely.\n- The injected template is evaluated at runtime and resolves to whatever the Jinja function returns (`event()` gives the full event object, `environment()` gives tenant metadata, plus other available functions per the Dynatrace Workflows documentation).\n- The resolved output is delivered to a destination the attacker controls (the `channel` parameter is also templated, and is even more dangerous because it is interpolated inside an existing `{{ \"...\" }}` expression context).\n- The workflow persists in the tenant after the MCP session, operator credentials, and MCP server are gone.\n- With `isPrivate=false` (the default), the workflow is visible to all tenant users.\n\nExploitation requires either prompt injection (an LLM under the MCP server\u0027s control is reading attacker-controlled data and follows an injection that calls `create_workflow_for_notification` with the malicious arguments) or operator carelessness (the approval prompt shows the raw team name literal and the operator clicks Approve without recognising `{{ event() }}` as a template fragment). Both are realistic in practice.",
  "id": "GHSA-xrmj-5g4g-8987",
  "modified": "2026-07-31T16:01:07Z",
  "published": "2026-07-31T16:01:07Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/dynatrace-oss/dynatrace-mcp/security/advisories/GHSA-xrmj-5g4g-8987"
    },
    {
      "type": "WEB",
      "url": "https://github.com/dynatrace-oss/dynatrace-mcp/pull/547"
    },
    {
      "type": "WEB",
      "url": "https://github.com/dynatrace-oss/dynatrace-mcp/commit/64dfcb1095823d0fb43013635446288a6ff60e29"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/dynatrace-oss/dynatrace-mcp"
    },
    {
      "type": "WEB",
      "url": "https://github.com/dynatrace-oss/dynatrace-mcp/releases/tag/v2.0.0"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:H/PR:N/UI:R/S:U/C:L/I:L/A:N",
      "type": "CVSS_V3"
    }
  ],
  "summary": "@dynatrace-oss/dynatrace-mcp-server has a workflow template injection via create_workflow_for_notification"
}

Mitigation
Architecture and Design

Choose a template engine that offers a sandbox or restricted mode, or at least limits the power of any available expressions, function calls, or commands.

Mitigation
Implementation

Use the template engine's sandbox or restricted mode, if available.

No CAPEC attack patterns related to this CWE.