reflect

Inventory user feedback and constraints from conversation context before proceeding.

15|6|Updated Apr 20, 2026
One-click install
npx skills add https://github.com/phnx-labs/agents-cli --skill reflect-phnx-labs
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: reflect
Source: https://github.com/phnx-labs/agents-cli/tree/main/.agents/skills/reflect
Command: npx skills add https://github.com/phnx-labs/agents-cli --skill reflect-phnx-labs

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill prevents AI drift by forcing the model to explicitly inventory and incorporate all previous user feedback, corrections, and constraints before generating new responses.

Core Features & Use Cases

  • Feedback Inventory: Systematically extracts and lists all user corrections, rejections, and constraints from the conversation history.
  • Pattern Identification: Identifies the underlying thread or recurring issue causing iterative failures.
  • Use Case: When an AI repeatedly fails to follow a specific formatting constraint or keeps ignoring a stylistic preference, this skill forces a comprehensive review of the conversation to realign the model's output with user expectations.

Quick Start

Invoke the reflect skill to review all previous feedback and constraints before generating the next iteration of your current task.

Frequently Asked Questions about reflect

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

FAQPage Schema
How do I prevent AI from ignoring constraints and previous feedback in a conversation?

To prevent AI from ignoring constraints, you systematically extract and list all user corrections from the conversation history before generating new responses. This process realigns the model's output with user expectations and prevents iterative failures caused by AI drift.

Why does AI drift happen during iterative refinement and how to fix it?

AI drift happens when the model loses track of accumulated constraints over long interactions. You fix this by requiring a formal feedback inventory, identifying recurring failure patterns, and establishing a revised execution strategy before proceeding with subsequent tasks.

What is the best way to ensure global constraint satisfaction across multiple tasks?

The best way to ensure global constraint satisfaction is to synthesize all user-provided feedback and corrections from the current conversation context. By explicitly inventorying this data, you establish a revised execution strategy that satisfies all constraints before proceeding.

Can I use this feedback recall approach for stylistic preference corrections?

Yes, you can use this approach for stylistic preference corrections. When an AI repeatedly fails to follow formatting constraints or stylistic preferences, a comprehensive conversation review forces the model to realign its output with your specific expectations.

How do I identify patterns causing iterative failures in AI outputs?

To identify patterns causing iterative failures, you systematically extract and list all user corrections and rejections from conversation history. This feedback inventory reveals the underlying thread or recurring issue, allowing you to establish a revised execution strategy.