← ClaudeAtlas

rememberlisted

Save one durable lesson, rule or preference; use when the user states a standing preference or an approach fails in a way worth never repeating.
mubit-ai/plugins · ★ 3 · AI & Automation · score 76
Install: claude install-skill mubit-ai/plugins
Routine work is captured automatically — do not use this for "I read a file" or "I ran the tests". The capture hooks already write tool activity, prompts, and turn outcomes into this run without being asked. Using the explicit verb for that fills the store with duplicates of what is already there and buries the entries that matter. Use it for knowledge that should outlive this session. Reach for it when: - the user states a standing preference ("always run the suite before you commit"); - you discover a constraint that is not visible in the code ("the operator's reconcile loop requires the CRD applied before the StatefulSet, or it wedges"); - an approach fails in a way worth never repeating, and you know *why* it failed. Call `mcp__mubit__mubit_learned`. Pick a template, which sets `lesson_type` and `lesson_scope` for you: | Template | Use for | type / scope | | --- | --- | --- | | `CODING_RULE` | lint rules, naming, style constraints | rule / global | | `DEBUG_SUCCESS` | a debugging approach that worked | success / session | | `DEBUG_FAILURE` | an approach that failed, and why | failure / session | | `PREFERENCE` | how this user wants things done | preference / global | | `ARCHITECTURE_INSIGHT` | system quirks, dependency behaviour | observation / global | | `BUILD_CONFIG` | build/deploy settings that work | rule / global | | `API_PATTERN` | SDK quirks, integration notes | observation / session | | `TEST_STRATEGY` | test approaches that proved effective | success / glo