Capture behavioral friction as structured observations in JSON files.
npx skills add https://github.com/Miosa-osa/canopy --skill remember-miosa-osa
Please help me install this Agent Skill. Skill: remember Source: https://github.com/Miosa-osa/canopy/tree/main/library/skills/learning/remember Command: npx skills add https://github.com/Miosa-osa/canopy --skill remember-miosa-osa
Capture behavioral friction as structured observations to support continuous improvement of AI systems. When users correct behavior, when patterns emerge, or when conversations reveal gaps, store actionable lessons for later synthesis.
Tell /remember a lesson learned to capture the observation for system improvement.
High-intent search queries and answers about installing and using this skill.
Capture behavioral friction by logging structured observations from explicit user feedback, contextual corrections, or mined transcripts. These observations are stored as JSON files with category, severity, and confidence metadata to drive system improvements.
Extracting lessons from transcripts uses mining mode to scan for recurring frictions and shortcuts. The system surfaces these patterns and stores them as structured observations for later synthesis, ensuring no behavioral feedback is lost.
Deduplication is handled automatically within the formal observation model. When logging corrections, the system checks existing stored observations to prevent redundant entries, ensuring your feedback logs remain clean and actionable.
Yes, contextual mode detects and extracts corrections directly from ongoing conversations. It automatically captures behavioral friction as structured observations without requiring explicit user commands to log the feedback.
Structured observations are stored as JSON files at ops/observations/{category}/{timestamp}-{slug}.json. This format includes formal metadata like category, severity, confidence, and deduplication status to support later analysis.
Stored observations are synthesized into actionable improvements via the /rethink command. This later synthesis step aggregates the captured feedback, analyzing the structured JSON files to identify and resolve systemic behavioral friction.
outfitter-devAnalyze conversation transcripts to detect patterns, signals, and behaviors.
kinnerkarmanishCapture behavioral friction and lessons as structured observations for the rethink loop.
micherraAnalyze codebase patterns and review history to generate improvement proposals.
MantisWareCapture and organize lessons and behavioral corrections as structured observations.
