Analyze an agent's session behavior to identify process errors and knowledge gaps.
npx skills add https://github.com/korchasa/flowai-plugins --skill reflect-korchasa
Please help me install this Agent Skill. Skill: reflect Source: https://github.com/korchasa/flowai-plugins/tree/main/plugins/flowai/skills/reflect Command: npx skills add https://github.com/korchasa/flowai-plugins --skill reflect-korchasa
The reflect Skill addresses the need for in-depth analysis of an agent's behavior and performance during a session, identifying errors, inefficiencies, and missed opportunities for improvement.
Use the reflect skill to analyze the last session for insights on process, technical decisions, and context usage.
High-intent search queries and answers about installing and using this skill.
To analyze an agent's technical decisions, you review the session history to evaluate the soundness of technical choices made during the task. This identifies technical missteps and logical errors to improve future performance.
Agent session analysis is the comprehensive review of an agent's process, context usage, and knowledge gaps. You need it after a task to identify process errors, inefficiencies, and missed opportunities for improvement.
You detect process errors and knowledge gaps by evaluating the session history against the expected task workflow. This identifies context-related inefficiencies and technical missteps where the agent lacked necessary information.
Yes, reviewing agent session history identifies repetitive tasks that can be automated to increase efficiency. By analyzing the process workflow, you can pinpoint areas where manual or redundant steps should be automated.
Yes, evaluating agent context usage requires access to session history and task management tools. These tools provide the necessary task data and session metrics to properly assess context-related inefficiencies.
The best way to review technical missteps is to perform a comprehensive session analysis that evaluates the agent's process and technical decisions. This method isolates logical errors and context inefficiencies effectively.
CarbeneAIAuto-loads core PAI identity, response format, and routing triggers at session start.
cropsggScan .commandcode/learnings, prune outdated entries, and export summaries to EXPORT_YYYYMMDD.md.
knowledgeislandsSummarize live session details and route outstanding items and learnings.
kynopticAnalyze stack traces and error logs to identify root causes and propose fixes.
