67 lines
1.9 KiB
Markdown
67 lines
1.9 KiB
Markdown
# Didactopus FAQ: Artifact Lifecycle and Knowledge Reuse
|
|
|
|
## Why keep artifacts after rendering?
|
|
|
|
Artifacts are evidence of learning trajectories, pack structure, and interpretation.
|
|
They support:
|
|
- learner reflection
|
|
- mentor review
|
|
- debugging AI learners
|
|
- presentation and publication
|
|
|
|
## Why do retention policies matter?
|
|
|
|
Not every artifact should be stored forever. Some are transient debugging outputs;
|
|
others are durable portfolio items or research artifacts.
|
|
|
|
Retention policy support lets deployments distinguish:
|
|
- short-lived temporary outputs
|
|
- retained educational outputs
|
|
- archival artifacts worth preserving
|
|
|
|
## How can learner knowledge be used outside Didactopus?
|
|
|
|
A learner's activity can be exported into structured forms that support:
|
|
- revised or expanded domain packs
|
|
- lesson plans and conventional curriculum products
|
|
- AI skill definitions or prompts
|
|
- mentor-facing notes about misconceptions and discoveries
|
|
|
|
## Can learners improve domain packs?
|
|
|
|
Yes. Learners sometimes notice:
|
|
- confusing sequence order
|
|
- hidden prerequisites
|
|
- missing examples
|
|
- better analogies
|
|
- edge cases mentors overlooked
|
|
|
|
Didactopus should capture these as improvement suggestions rather than losing them.
|
|
|
|
## How could this support agentic AI skills?
|
|
|
|
A learner knowledge export can be mapped into:
|
|
- scope and goals
|
|
- prerequisite structure
|
|
- canonical examples
|
|
- failure modes
|
|
- evaluation checks
|
|
- recommended actions
|
|
|
|
That makes it a plausible source for building agent skills or skill-like bundles.
|
|
|
|
## How could this support traditional curriculum products?
|
|
|
|
Knowledge export can seed:
|
|
- lesson outlines
|
|
- exercise sets
|
|
- study guides
|
|
- formative assessment prompts
|
|
- instructor notes
|
|
- capstone project ideas
|
|
|
|
## Is exported learner knowledge treated as automatically correct?
|
|
|
|
No. Exported learner knowledge should be treated as candidate structured knowledge.
|
|
It is useful, but it still needs review, validation, and provenance tracking.
|