Capture friction and user corrections as methodology notes to refine AI behavior.
npx skills add https://github.com/zby/llm-do --skill remember-zby
Please help me install this Agent Skill. Skill: remember Source: https://github.com/zby/llm-do/tree/main/arscontexta/skills/arscontexta-remember Command: npx skills add https://github.com/zby/llm-do --skill remember-zby
This Skill helps capture and formalize learnings from interactions, turning friction points and corrections into actionable methodology notes that improve AI behavior over time.
/remember "Don't process personal notes like research" to capture that correction. Later, if you notice a pattern of similar mistakes, you can use /remember --mine-sessions to scan past interactions for these issues and generate new methodology notes.Use the remember skill to capture the instruction "always check for duplicates before creating".
High-intent search queries and answers about installing and using this skill.
To capture AI corrections, you can provide explicit instructions or analyze recent conversation history for implicit adjustments. This turns friction points into actionable methodology notes that refine AI behavior over time.
Session mining scans past session transcripts to identify recurring patterns of friction or undocumented decisions. It generates new methodology notes from these patterns to continuously evolve system specifications.
You document implicit user adjustments by using contextual review to analyze recent conversation history. This identifies undocumented corrections and integrates them into a methodology learning loop to refine AI behavior.
No external dependencies are required to run the methodology learning loop. It operates independently to capture friction and evolve system specifications without needing additional software engineering frameworks.
Use explicit capture for direct instructions like specific corrections, and use contextual review when analyzing recent conversation history for implicit adjustments. Both generate methodology notes to refine AI behavior.
agenticnotetakingCapture friction points as structured markdown methodology notes from conversations.
wojonsGuide AI collaboration through a 10-step software development methodology.
tachyon-beepRoute LLM tasks to specialized skills via a decision tree.
swarm-ai-safetyRecord methodology friction and process corrections into the SWARM vault.