Create an assessmentTest XML document with a unique identifier and title
Add a testPart with navigationMode (linear or nonlinear) and submissionMode (individual or simultaneous)
Within the testPart, define one or more assessmentSections, each with a selection and ordering element if randomization is desired
Add assessmentItemRef elements within each section, each referencing an item by href path and carrying a weight attribute
Set a timeLimits element at the testPart or test level with maxTime expressed in seconds
Include an outcomeDeclaration at the test level for SCORE and a testFeedback element for pass/fail messaging if needed
Known gotchas
timeLimits at the test level and at the testPart level interact; a testPart time limit is enforced independently of the overall test limit
Delivery engines may not support all navigationMode and submissionMode combinations; verify the target platform's conformance profile before authoring
Item weights in assessmentItemRef must sum to a meaningful total; an uncalibrated weight of 1.0 on all items effectively gives equal weight, which may not match intended scoring
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