Analyze conversation feedback to propose patch-level updates to existing SKILL.md files.
npx skills add https://github.com/materkey/cc-plugins --skill reflect-materkey
Please help me install this Agent Skill. Skill: reflect Source: https://github.com/materkey/cc-plugins/tree/main/plugins/reflect/skills/reflect Command: npx skills add https://github.com/materkey/cc-plugins --skill reflect-materkey
This Skill helps you capture repeated user corrections, approvals, and preferences from the current conversation and convert them into concrete improvements for an existing Skill. It reduces the chance of repeating the same mistakes across future sessions by updating the Skill instructions directly.
Ask the assistant to run the reflect skill on the current conversation and propose approved edits for any existing skill that the user corrected.
High-intent search queries and answers about installing and using this skill.
To update skill instructions from user corrections, analyze the active conversation to detect repeated feedback and skill failures, then map those signals to existing SKILL.md files and apply targeted patch-level edits to improve future behavior.
Capturing workflow preferences for prompt improvement involves scanning the current session for confirmed patterns and user approvals, filtering the findings to changes that belong inside an existing skill definition, and directly editing the approved instructions.
Conversation analysis detects skill failures by reading conversation evidence to identify recurring user corrections, mapping those signals to specific problems in a skill's steps, criteria, or formatting, and separating patch-level issues from architecture-level concerns.
You need to escalate a skill failure to an architecture-level workflow when the issue is larger than a patch, requiring a new system component rather than modifying existing steps, criteria, formatting, or behavioral instructions in the current SKILL.md file.
Yes, conversation signal analysis works for fixing output formatting rules by identifying repeated user corrections regarding formatting within the session and applying those findings as targeted edits to the relevant existing skill definitions.
Yes, you can use conversation reflection to correct tool selection choices by identifying where the user repeatedly corrected the assistant's tool choices, then applying those workflow corrections directly into the existing skill's behavioral instructions.
Largo2z9Encode operator corrections as durable Hard Rules in SKILL.md files.
easonlaoPlan and execute a fix-only day-one patch with rollback plan and QA gating.
lawvableAnalyze conversations to extract actionable improvement instructions for other skills.
linenoizeAudit workflow conversations for mandatory skill constraint compliance and violations.
