retro-analysis

Analyze agent workflow logs and comments to identify issues and recommend improvements.

1|16|Updated Apr 23, 2026
One-click install
npx skills add https://github.com/konflux-ci/.fullsend --skill retro-analysis-konflux-ci
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: retro-analysis
Source: https://github.com/konflux-ci/.fullsend/tree/main/skills/retro-analysis
Command: npx skills add https://github.com/konflux-ci/.fullsend --skill retro-analysis-konflux-ci

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill helps teams perform retrospectives on agent workflows by tracing run sequences, exploring context, and identifying areas for improvement, facilitating continuous process enhancement.

Core Features & Use Cases

  • Workflow Tracing: Reconstructs the sequence of agent runs from PRs or issues to visualize the workflow graph.
  • Log Analysis: Downloads and reads reasoning traces and logs to understand agent decision-making and errors.
  • Comment Categorization: Analyzes PR comments to distinguish between agent feedback, human input, and test results.
  • Improvement Proposals: Synthesizes findings into actionable suggestions to refine agent behavior and configuration.
  • Use Case: A team reviewing a failed agent run uses this Skill to pinpoint divergence points and formulate targeted fixes.

Quick Start

Use the retro-analysis skill to analyze the latest agent workflow run related to issue #123 in the repository.

Frequently Asked Questions about retro-analysis

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

FAQPage Schema
How do I analyze agent workflow logs to identify process issues?

Analyzing agent workflow logs involves reconstructing run sequences from PRs or issues, reading reasoning traces, and categorizing comments to pinpoint decision errors and propose targeted process improvements.

What is agent decision tracing and how does it support team retrospectives?

Agent decision tracing is the process of downloading and reading reasoning traces from agent runs to understand decision-making. It supports retrospectives by visualizing workflow graphs and identifying areas for workflow enhancement.

How do I trace agent run histories from a specific issue or pull request?

Tracing agent run histories from a specific issue or PR reconstructs the sequence of agent runs to visualize the workflow graph, allowing you to analyze context and pinpoint divergence points for targeted fixes.

Can I categorize PR comments to distinguish between human input and agent feedback?

Categorizing PR comments distinguishes between agent feedback, human input, and test results by analyzing PR discussions, helping teams understand interaction patterns and synthesize findings into actionable configuration suggestions.

What is the best way to propose improvements for a failed agent run?

Proposing improvements for a failed agent run requires synthesizing findings from workflow tracing and log analysis to pinpoint divergence points, resulting in targeted fixes for agent behavior and configuration.

Does workflow analysis work without dependencies for continuous process enhancement?

Workflow analysis for continuous process enhancement operates without external dependencies, directly reading agent run histories, reasoning traces, and PR discussions to formulate targeted fixes and improvement proposals.