A pre-test and post-test check how your judgment changes on when to use an agent, planning, tool design, read/write separation, semantic layers, security, verification, prompt injection, evaluation, observability, failure recovery, and human-in-the-loop approval
Assessment flow
DS 302 follows the platform’s assessment pattern: take the pre-test, learn and practice, then take the post-test. The pre-test and post-test are parallel but not identical forms, so learning gains can be observed.
1Pre-test Completed before the course starts, to establish your current judgment. The pre-test has no passing threshold.
2Learn Complete all ten lessons across Module 1 and Module 2.
3Practice Complete eight practice tasks.
4Post-test Taken after finishing the course; the passing threshold is 80%.
Assessment content
Each test has 12 single-select questions covering twelve constructs: agent need, plan, tools, read/write separation, semantic layer, security, verify, prompt injection, evaluation, observability, failure recovery, and human-in-the-loop approval. Questions test judgment, not tool-operation memorization.
Question distribution per test
Construct
Questions
agent need
1
plan
1
tools
1
read/write separation
1
semantic layer
1
security
1
verify
1
prompt injection
1
evaluation
1
observability
1
failure recovery
1
human-in-the-loop
1
How scoring works
Scoring happens on the server. Correct answers and explanations never appear in public pages, downloadable files, or any response before submission. After submission, explanations are shown only for questions you missed.
Info
Starting the assessment requires a cohort code from your instructor and a learner code. Use your learner code, not your real name.
