@xaccefy/learnmax
Research sources. Build lessons. Check what stuck — in one loop.
Package details
Install @xaccefy/learnmax from npm and Pi will load the resources declared by the package manifest.
$ pi install npm:@xaccefy/learnmax- Package
@xaccefy/learnmax- Version
1.3.1- Published
- Jul 26, 2026
- Downloads
- 813/mo · 365/wk
- Author
- xaccefy
- License
- MIT
- Types
- skill
- Size
- 29 KB
- Dependencies
- 0 dependencies · 0 peers
Pi manifest JSON
{
"skills": [
"./skills"
]
}Security note
Pi packages can execute code and influence agent behavior. Review the source before installing third-party packages.
README
learnmax
Research sources. Build lessons. Check what stuck — in one loop.
/learn "memory layout" → research sources, build lessons
/quiz "memory layout" → self-check
/test "memory layout" → spaced review, scores, feeds weak spots back
/learn builds everything in one call. /quiz and /test run separately, whenever you want to check yourself.
Install
pi install npm:@xaccefy/learnmax
Then from any directory you want to use as a learning workspace:
/learn "memory layout"
/quiz "memory layout"
/test "memory layout"
How it works
Three commands, one loop:
| Command | What it does |
|---|---|
/learn |
Researches validated sources, builds Markdown lessons, glossary, and literature review under <topic>/. Picks up the weak-item backlog first if it exists. |
/quiz |
Self-check over lessons and glossary. Immediate feedback, writes nothing. |
/test |
Spaced retrieval review. Scores known / shaky / unknown, writes relearn/, and rewrites the weak-item backlog. |
The weak-item loop
/test rewrites relearn/weak-items.md every run. Items that stay weak reset their streak. Items you recall move through 1/3/7-day spacing intervals. Recalled three times in a row = retired.
Next /learn grabs the backlog first — weak spots drive new lessons automatically, and mastered items leave the loop instead of being re-taught forever.
Workspace layout
Workspace root:
TOPICS.md # registry of which topics exist
relearn/ # /test output (gitignored)
<date>.md # recall scores from /test
weak-items.md # backlog — /learn reads this first
<topic>/ # one directory per topic
MISSION.md # why you're learning this
GLOSSARY.md # canonical terms, linked to their lesson
literature/ # source maps — consensus, tensions, gaps
lessons/
NNNN-slug.md # one lesson per file, Obsidian-ready
Commit everything under <topic>/. relearn/ is temporary — /learn ignores it.
Format specs
All artifact formats are defined in skills/learn/FORMATS.md.
Validation
After building, /learn runs learnmax-validate — checks every topic dir against the format specs and fails on format drift. Run it manually with:
node <pkg>/scripts/validate-artifacts.mjs
License
MIT.