Adobe Customer Journey Analytics
Run cross-channel reports and build segments in Adobe Customer Journey Analytics.
Look up US K-12 standards, break them into learning components, trace prerequisites.
Opens Agent Studio, where connecting is one click. The connector URL below works in any MCP client.
Paste it into any MCP client. Setup docs
The Learning Commons MCP server gives Claude, ChatGPT and any MCP agent authoritative US K-12 academic standards. Three read-only tools resolve a standard code into its official statement, break it into granular learning components, and trace prerequisite progressions. It reaches a public reference graph, never a student record, and the data is CC BY 4.0.
Verified connector
Listed by Anthropic as a partner connector in its Connectors Directory.
Connection checked by Agentman on .
Anthropic states this reflects the level of review a connector received, not a security audit.
Tool names from Anthropic's directory listing. This server requires sign-in, so we could not read tool descriptions or parameter schemas.
docs.learningcommons.org/knowledge-graph/using-knowledge-graph/claude-connector, which returned 404 on 2026-08-23. The server and the documentation are both alive — the working page is .../agent-plugins, and the docs host serves a 2,287-word llms.txt index and a 69,210-word llms-full.txt. A dead doc URL is not a dead connector, and this entry has been misclassified as one before.find_standards_progression_from_standard does not say so.isCurrent=false rather than removed.caseIdentifierUUID, not a human-readable code, so a question about a standard code has to resolve the statement first. Chain the tools; do not expect either to work from a bare code.attributionStatement field arrives with the response, but nothing forces it to survive into whatever the agent writes.readOnlyHint or destructiveHint is reported here. Parameter detail comes from the vendor's documented input table, not from the server.No. All three tools take a standard code or a CASE UUID and return published reference data about that standard. Nothing in the surface accepts a student name, roster, score or submission, and no tool writes anywhere. Learning Commons publishes no student-data commitment for the connector because the connector never reaches a system that holds any.
Not in Claude. Learning Commons states the connector is free and available to all Claude users, and its Claude setup steps ask only for a Claude account. The raw endpoint does require a key — an anonymous request returned HTTP 401 with a www-authenticate Key challenge on 2026-08-23 — so direct MCP clients need a Learning Commons Platform key.
The model generates them; the connector only supplies the facts. All three tools read. Learning Commons' entire Knowledge Graph REST API is 31 operations and every one is a GET, with no POST, PUT, PATCH or DELETE anywhere. Anthropic records the permissions as Read. The word Generate in the listing describes Claude's output, not a tool.
Standards lookup covers all 50 states; the other two tools cover much less. Math learning components span 42 jurisdictions across K-12, but ELA learning components are K-2 only across 13 jurisdictions. Progressions come solely from the Math Coherence Map, so an ELA prerequisite question has no data behind it.
Yes. Knowledge Graph data is licensed CC BY 4.0, which makes attribution a licence condition rather than a courtesy. Learning Commons documents that standards responses carry the source attribution the licence requires, and its API returns an attributionStatement field. Code is MIT; the progressions are CC0 from Student Achievement Partners.
Learning Commons labels it early access. Its MCP server page carries a badge saying the functionality is actively evolving and that breaking changes may occur as it expands capabilities and improves accuracy. The connector itself is listed by Anthropic as a partner connector, but the vendor's own stability warning applies to the server behind it.
Yes, on the underlying API. Learning Commons documents 2 requests per second with burst capacity up to 10, HTTP 429 on breach, and advises exponential backoff. It invites teams to email support to request an adjustment. The published figures describe the Knowledge Graph API; the vendor states no separate MCP limit.
No, and Learning Commons forbids it. Its terms of use state the resources are provided for exploratory and research use and are not intended for high-stakes educational decisions such as formal student evaluation, instructional placement, or grading. It also states human judgement is expected when interpreting results.
.../using-knowledge-graph/claude-connector, returned 404 the same day. · retrieved 2026-08-23llms.txt and llms-full.txt (retrieved 2026-08-23). docs.learningcommons.org/robots.txt carries Content-Signal: ai-train=yes, search=yes, ai-input=yes — synthesis expressly permitted by the publisher. · retrieved 2026-08-23initialize to https://kg.mcp.learningcommons.org/mcp returned 401 with www-authenticate: Key; four RFC 9728 probes under /.well-known/oauth-protected-resource, plus /.well-known/oauth-authorization-server and the host root, all returned matching Kong 404s · retrieved 2026-08-23Connect Learning Commons once and your agents call these tools on their own: on a schedule, in a workflow, with nobody at the keyboard.