retrospective

Capture and classify user friction signals during Claude Code sessions.

1|Updated Apr 13, 2026
One-click install
npx skills add https://github.com/prince-khanna/claude-code-plugins --skill retrospective-prince-khanna
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: retrospective
Source: https://github.com/prince-khanna/claude-code-plugins/tree/main/personal-assistant/skills/retrospective
Command: npx skills add https://github.com/prince-khanna/claude-code-plugins --skill retrospective-prince-khanna

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The retrospective skill captures moments of user friction during a Claude Code session, logs them, and routes the insights to the appropriate context files so Imli can learn and improve without manual bookkeeping.

Core Features & Use Cases

  • Real‑time friction capture: silently notes corrections, redo requests, frustration signals, and repeated back‑and‑forth while the session runs.
  • Interactive retrospective: on /retrospective it scans the whole conversation, classifies each friction moment, prioritises up to five findings, and writes them to the correct context files (rules, preferences, workflows, project memory, improvements).
  • Safety checks: avoids duplicates, contradictions, and oversized files; verifies proposals and updates the journal.

Quick Start

Invoke /retrospective at the end of a session to have Imli summarise and store any friction you experienced.

Frequently Asked Questions about retrospective

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

FAQPage Schema
How do I capture user friction and correction moments during a Claude Code session?

User friction signals and correction moments are captured silently during a Claude Code session by noting corrections, redo requests, and repeated back-and-forth interactions to identify areas for improvement.

How do I run a retrospective to classify and prioritize session friction?

Run a retrospective by scanning the whole conversation to classify each friction moment, prioritizing up to five findings, and routing them to appropriate context files like rules, preferences, or workflows.

How does the retrospective skill route findings to context files safely?

The retrospective skill routes findings safely by verifying proposals, avoiding duplicates and contradictions, preventing oversized files, writing confirmed entries to the correct context files, and updating the session journal.

What is the best way to automate session improvement without manual bookkeeping?

Automating session improvement without manual bookkeeping is achieved by capturing real-time friction signals, classifying them during a retrospective, and automatically writing the insights to project memory and rules files.

Does the retrospective skill require any dependencies to run?

No dependencies are required to run the retrospective skill. It operates independently to scan conversations, classify friction signals, and update context files within the development environment.

When should I avoid using an automated retrospective for context file management?

You should avoid relying solely on an automated retrospective when context files are already oversized or prone to contradictions, as the skill prioritizes safety checks to prevent duplicate or conflicting entries.