bee:dev-feedback-loop

Analyze development cycle metrics by automating the collection of assertiveness scores, gate iterations, and prompt quality from task state files and agent output logs to identify recurring failure patterns and generate actionable improvement suggestions.

Root cause analysis for low assertiveness scores performs deep dives into tasks with excessive iterations, evaluating agent prompts for effectiveness and detecting systemic issues across development gates to highlight specific process bottlenecks.

Yes, you can evaluate AI agent prompt quality automatically by analyzing agent output logs to assess effectiveness, identifying uncovered criteria in gates like testing, and suggesting specific prompt adjustments to improve coverage.

The best way to collect assertiveness scores and iteration counts is by accessing task state files and agent output logs to automate metric gathering across each development gate, enabling comprehensive cycle analysis and feedback reporting.

Yes, you need access to task state files and agent output logs as prerequisites for continuous improvement analysis, because these files provide the raw cycle metrics and prompt data required to identify failure patterns and generate feedback.

Your testing gate requires more iterations due to systemic prompt deficiencies in the QA Analyst agent, which continuous improvement analysis identifies by evaluating prompt quality and suggesting specific adjustments to improve test coverage.