From qe-framework
Manages cross-session project memory by storing conventions, gotchas, decisions, and patterns with TTL-based expiry. Use /Qmemory to add, list, or prune entries.
How this skill is triggered — by the user, by Claude, or both
Slash command
/qe-framework:QmemoryThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Manages `.qe/project-memory.json` for persistent cross-session knowledge. Stores conventions, gotchas, decisions, and patterns with TTL-based expiry.
Manages .qe/project-memory.json for persistent cross-session knowledge. Stores conventions, gotchas, decisions, and patterns with TTL-based expiry.
/Qmemory add "content" [--type TYPE] [--priority PRIORITY] [--tags tag1,tag2]Add a memory entry.
convention | gotcha | decision | pattern (default: convention)permanent | high | normal | low (default: normal)TTL by priority:
| Priority | TTL |
|---|---|
| permanent | never expires |
| high | 30 days |
| normal | 7 days |
| low | 1 day |
/Qmemory list [--type TYPE] [--tag TAG]List active (non-expired) memories. Optionally filter by type or tag. Display as a table with id, type, priority, content, and expiry.
/Qmemory pruneRemove expired entries. Report how many were pruned.
/Qmemory clearClear all entries. Confirm with the user before executing.
Extract the subcommand (add, list, prune, clear) and any flags from the user input.
Use hooks/scripts/lib/project-memory.mjs functions:
addMemory(cwd, content, type, { priority, source: 'user', tags }). Print the created entry.getActiveMemories(cwd), or getMemoriesByType/getMemoriesByTag if filtered. Display as a formatted table.pruneExpired(cwd). Report count.clearAll(cwd).Show a concise confirmation or table. No verbose explanations.
.qe/analysis/ files (those are for codebase structure)Project memory: {count} active entries ({permanent} permanent, {expiring} expiring)
npx claudepluginhub inho-team/qe-framework --plugin qe-frameworkCreates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.