generated: '2026-08-13' method: searched source: https://experienceleague.adobe.com/llms.txt (Adobe's own published llms.txt, "MCP Server Access" section), https://github.com/adobe/skills (plugins/adobe-analytics/.mcp.json and plugins/adobe-cja/.mcp.json), and live JSON-RPC probes of every endpoint on 2026-08-13 status: published description: >- Adobe ships REMOTE, Adobe-hosted MCP servers for Experience Cloud. There are three distinct hosts: aa-mcp.adobe.io (Adobe Analytics), cja-mcp.adobe.io (Customer Journey Analytics), and mcp.adobeaemcloud.com (Adobe Experience Manager as a Cloud Service, seven server paths). Every endpoint is a live HTTPS URL an MCP client POSTs to — none of them is a stdio package a human has to install first. All are OAuth-gated behind Adobe IMS, so tools/list returns 401 anonymously and the live inputSchemas require an authenticated introspection. deployment: mode: remote endpoint: https://aa-mcp.adobe.io/mcp auth: oauth verified: probed note: >- Nine remote endpoints across three hosts are enumerated under endpoints[] below; aa-mcp.adobe.io/mcp is named here as the primary Experience Cloud analytics surface. No stdio/npx package exists — registry.npmjs.org returns 404 for @adobe/aep-mcp and @adobe/experience-platform-mcp. Adobe distributes the CLIENT-side configuration (the .mcp.json template plus agent skills) through the github.com/adobe/skills plugin marketplace, but the servers themselves are hosted by Adobe. endpoints: - name: adobe-analytics product: Adobe Analytics url: https://aa-mcp.adobe.io/mcp transport: http auth: oauth probe: method: 'POST {"jsonrpc":"2.0","id":1,"method":"tools/list"}' http_status: 401 body: Missing or empty Authorization token www_authenticate: Bearer realm="OAuth", resource_metadata="https://aa-mcp.adobe.io/.well-known/oauth-protected-resource" checked: '2026-08-13' oauth: protected_resource: https://aa-mcp.adobe.io/.well-known/oauth-protected-resource authorization_server: https://aa-mcp.adobe.io/.well-known/oauth-authorization-server authorization_endpoint: https://ims-na1.adobelogin.com/ims/authorize/v2 token_endpoint: https://ims-na1.adobelogin.com/ims/token/v3 registration_endpoint: https://aa-mcp.adobe.io/register scopes: - openid - AdobeID - additional_info.projectedProductContext pkce: S256 dynamic_client_registration: true additional_headers: - Authorization - x-gw-ims-org-id - x-gw-ims-user-id source: https://github.com/adobe/skills/blob/main/plugins/adobe-analytics/.mcp.json - name: adobe-cja product: Adobe Customer Journey Analytics url: https://cja-mcp.adobe.io/mcp transport: http auth: oauth probe: method: 'POST {"jsonrpc":"2.0","id":1,"method":"tools/list"}' http_status: 401 body: Missing or empty Authorization token www_authenticate: Bearer realm="OAuth", resource_metadata="https://cja-mcp.adobe.io/.well-known/oauth-protected-resource" checked: '2026-08-13' oauth: protected_resource: https://cja-mcp.adobe.io/.well-known/oauth-protected-resource/mcp authorization_server: https://cja-mcp.adobe.io/.well-known/oauth-authorization-server authorization_endpoint: https://ims-na1.adobelogin.com/ims/authorize/v2 token_endpoint: https://ims-na1.adobelogin.com/ims/token/v3 registration_endpoint: https://cja-mcp.adobe.io/register scopes: - openid - AdobeID - additional_info.projectedProductContext pkce: S256 additional_headers: - Authorization - x-gw-ims-org-id - x-gw-ims-user-id - x-data-view-id source: https://github.com/adobe/skills/blob/main/plugins/adobe-cja/.mcp.json - name: aem-unified product: Adobe Experience Manager as a Cloud Service url: https://mcp.adobeaemcloud.com/adobe/mcp/aem transport: http auth: oauth probe: http_status: 401 body: 'JSON-RPC error 401 "Authorization header is required"' checked: '2026-08-13' source: https://experienceleague.adobe.com/en/docs/experience-manager-cloud-service/content/ai-in-aem/mcp-support/using-mcp-with-aem-as-a-cloud-service - name: aem-content product: Adobe Experience Manager as a Cloud Service url: https://mcp.adobeaemcloud.com/adobe/mcp/content transport: http auth: oauth scope: Create, read, update and delete pages and content fragments; asset import and asset search. Minimum AEM release 26309. probe: http_status: 401 www_authenticate: Bearer realm="oauth.adobeaemcloud.com", resource_metadata="https://mcp.adobeaemcloud.com/.well-known/oauth-protected-resource/adobe/mcp/content" checked: '2026-08-13' - name: aem-content-readonly product: Adobe Experience Manager as a Cloud Service url: https://mcp.adobeaemcloud.com/adobe/mcp/content-readonly transport: http auth: oauth scope: Read-only (Get, List/Search) for pages and content fragments, plus asset search. probe: http_status: 401 checked: '2026-08-13' - name: aem-cloudmanager product: Adobe Cloud Manager url: https://mcp.adobeaemcloud.com/adobe/mcp/cloudmanager transport: http auth: oauth scope: Manage Cloud Manager programs, environments, repositories and pipelines. probe: http_status: 401 checked: '2026-08-13' - name: aem-experience-governance product: Adobe Experience Manager as a Cloud Service url: https://mcp.adobeaemcloud.com/adobe/mcp/experience-governance transport: http auth: oauth scope: Evaluate content against brand integrity and compliance rules. probe: http_status: 401 checked: '2026-08-13' - name: aem-cloud-migration product: Adobe Experience Manager as a Cloud Service url: https://mcp.adobeaemcloud.com/adobe/mcp/cloud-migration transport: http auth: oauth scope: Retrieve Best Practices Analyzer (BPA) findings. probe: http_status: 401 checked: '2026-08-13' - name: aem-content-ai product: Adobe Experience Manager Content AI url: https://mcp.adobeaemcloud.com/adobe/mcp/content-ai transport: http auth: oauth scope: Keyword, semantic, hybrid and natural-language search over Content AI indexes. probe: http_status: 502 body: '{"title":"Bad Gateway","detail":"Internal routing error, please contact support."}' checked: '2026-08-13' note: Documented in the AEM MCP servers overview but returned 502 on probe. Recorded as documented-but-not-reachable, not as a verified endpoint. oauth_gateway: aem: issuer: https://oauth.adobeaemcloud.com authorization_endpoint: https://oauth.adobeaemcloud.com/oauth/authorize token_endpoint: https://oauth.adobeaemcloud.com/oauth/token registration_endpoint: https://oauth.adobeaemcloud.com/register scopes: - AdobeID - openid - read_organizations - aem.sites - aem.assets.author - aem.assets.delivery - aem.fragments.management - aem.folders - aem.repository - aem.contentai - aem.adobe.experimental - additional_info - additional_info.ownerOrg - additional_info.projectedProductContext pkce: S256 tools: note: >- tools/list returns 401 on every endpoint, so the live inputSchemas could not be read anonymously. The tool NAMES below are quoted verbatim from Adobe's own published agent skills in github.com/adobe/skills — they are what Adobe's skills invoke, not a guess. Parameter schemas require an authenticated introspection. source: https://github.com/adobe/skills (plugins/adobe-analytics/skills/*/SKILL.md, plugins/adobe-cja/skills/*/SKILL.md) adobe-analytics: - name: runReport description: Execute an Adobe Analytics report request. - name: findReportSuites description: Find/select the working report suite. - name: setSessionDefaults description: Set session context (reportSuiteId + globalCompanyId). - name: setDefaultReportSuite description: Set the default report suite for the session. - name: findMetrics description: Look up metrics available in the report suite. - name: findDimensions description: Look up dimensions available in the report suite. - name: findSegments description: Look up saved segments. - name: findCalculatedMetrics description: Look up calculated metrics. - name: searchDimensionItems description: Search item values inside a dimension. - name: describeSegment description: Describe a segment definition. - name: describeMetric description: Describe a metric definition. - name: upsertSegment description: Create or update a segment. - name: listComponentUsage description: Report component (dimension/metric/segment) usage. adobe-cja: - name: runReport description: Execute a Customer Journey Analytics report request. - name: findDataViews description: Find/select the working CJA data view. - name: setDefaultSessionDataViewId description: Set the default data view for the session. - name: findMetrics description: Look up metrics in the data view. - name: findDimensions description: Look up dimensions in the data view. - name: findSegments description: Look up filters/segments. - name: searchDimensionItems description: Search item values inside a dimension. - name: listComponentUsage description: Report component usage across the data view. aem_note: Named in the AEM MCP support documentation; the doc states tool availability evolves and that clients should discover tools at runtime via tools/list. aem: - name: get-aem-page-content - name: resolve_fragment_path - name: get_fragment - name: patch_fragment - name: list_models - name: get_model - name: create_fragment clients: documented_supported: - Anthropic Claude - OpenAI ChatGPT - Claude Code - Cursor - GitHub Copilot - Cline - Windsurf - OpenAI Codex - Kiro - Augment Code - Amazon Quick - Microsoft Copilot Studio source: https://experienceleague.adobe.com/en/docs/experience-manager-cloud-service/content/ai-in-aem/mcp-support/using-mcp-with-aem-as-a-cloud-service access_model: note: >- Quoted from Adobe's llms.txt — "MCP tools run under the authenticated user's identity and enforce the user's existing AEM permissions" and "MCP servers are designed for human-operated clients with interactive UX; the MCP Tools spec recommends a human in the loop to approve or deny tool invocations. Fully autonomous use should be treated as a separate compatibility tier."