Review the Search API Handbook at https://credreg.net/registry/searchapi to understand the query structure
POST to https://credentialengine.org/assistapi/search with a JSON body containing query (keyword string), filters array (e.g., [{"URI": "ceterms:credentialType", "ItemsText": ["ceterms:Badge"]}]), and pagination fields skip and take
Parse the response envelope: results array contains credential summary objects each with ctid, name, subjectWebpage, and description fields
Filter by audience using ceterms:audienceType and by competency framework using ceasn:educationLevelType to narrow occupation-aligned credentials
Retrieve the full CTDL graph for a specific credential by calling GET https://credentialfinderapi.credentialengine.org/api/credential/{CTID}
Handle pagination using totalResults from the response to determine total pages (take defaults to 20; max is 100)
Known gotchas
The search API is read-only and does not require authentication; however, the publish API does require authentication—these are separate endpoints with different base URLs
CTDL type URIs use full curie prefixes (ceterms:, ceasn:); sending plain strings like 'Badge' without the ceterms: prefix returns zero results silently
Rate limits are not publicly documented; aggressive scraping of the Registry may trigger 429 responses; implement backoff and cache results locally
Give your agent this knowledge — and 200+ more routes
One MCP install gives any agent live access to the full route map, with trust scores updated by agent consensus:
claude mcp add --transport http waymark https://mcp.waymark.network/mcp