skills.rest
codebytes
codebytesCOMMUNITY
Explore all skills by this author →
Website𝕏@chris_l_ayersGitHub

reflect

Extracts confidence-scored learning patterns from conversations to update agent history and team decisions.

Updated Jun 2, 2026
Search Tags:#learning capture#continuous improvement#pattern extraction#reflection#agent memory#team knowledge
One-click install
npx skills add https://github.com/codebytes/btt --skill reflect-codebytes
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: reflect
Source: https://github.com/codebytes/btt/tree/main/.squad/templates/skills/reflect
Command: npx skills add https://github.com/codebytes/btt --skill reflect-codebytes

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Teams repeatedly make the same mistakes because corrections and successful patterns from conversations are never captured. This Skill analyzes conversations for learning signals and proposes structured updates to persistent knowledge files so insights survive across sessions. ## Core Features & Use Cases - Confidence-Scored Signal Detection: Classifies learnings as HIGH (user corrections), MEDIUM (praise, edge cases), or LOW (accumulated preferences) with explicit thresholds for when to propose changes. - Routed Knowledge Persistence: Directs learnings to the right target: agent-specific history.md files, team-wide decisions inbox, or skill improvement recommendations. - Approval-Gated Updates: Always presents proposed changes in a structured summary and waits for user confirmation before writing to any file. - Use Case: After a user corrects an agent ("No, use the code search tools first, grep is too slow"), the Skill captures a HIGH-confidence constraint and appends it to the agent's history.md so the mistake is never repeated. ## Quick Start Ask the agent to reflect on this session and capture any learnings from corrections or feedback into the squad knowledge files.

Frequently Asked Questions about reflect

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I capture learnings from AI agent conversations?▼

Invoke the reflect Skill after user corrections, praise, or discovered edge cases. It scans the conversation for learning signals, scores them by confidence, and proposes structured updates to agent history or team decision files for your approval.

What triggers the reflect Skill to run?▼

HIGH priority triggers include user corrections like "no" or "wrong" and architectural insights. MEDIUM triggers include praise like "perfect" and tool preferences. LOW triggers include repeated usage patterns and session-end consolidation.

Where does the reflect Skill store captured learnings?▼

Agent-specific learnings append to .squad/agents/{agent}/history.md under a Learnings section. Team-wide patterns go to .squad/decisions/inbox/ as proposal files that a lead agent reviews before merging into decisions.md.

When should I not use conversation reflection?▼

Skip reflection for simple one-off questions with no reusable pattern, exploratory discussions without concrete decisions, or learnings already captured in history.md or decisions.md. Single LOW-confidence signals alone do not meet the proposal threshold.

Does reflect modify files without confirmation?▼

No. The Skill always displays proposed changes in a structured summary with source quotes and target files, then waits for explicit user approval (Y/n/edit) before writing anything to disk.

Related Skills

Explore More Software Engineering Skills→
githubnextgithubnext
5.1k

reflect

Extracts confidence-rated learning patterns from conversations to update agent history and team decision files.

Official
Intermediate
CommunityToolkitCommunityToolkit
622

reflect

Extracts confidence-rated learning patterns from conversations to update team knowledge files.

Official
Intermediate
paxtone-iopaxtone-io

sync

Synchronize the .kodo directory with Git or cloud storage.

Official
Intermediate
Soul-Brews-StudioSoul-Brews-Studio
116

retrospective

Capture retrospective notes and generate concise lessons-learned summaries via /rrr alias.

Official
Basic