ftm-retro

What problem does it solve?

Post-execution self-assessment is essential for improving reliability and throughput of ftm-executor plans. This skill scores execution across five evidence-based dimensions, surfaces bottlenecks with specifics, and builds a cumulative pattern library that makes each future execution smarter.

Core Features & Use Cases

• Automated scoring across five dimensions (wave parallelism, audit pass rate, Codex gate results, retry count, and execution smoothness) with explicit citations to data.
• Generates a formal retro report and persists it to the local state directory for later review and auditing.
• Pattern analysis and promotion: updates patterns.json to capture recurring lessons and guide future improvements.
• Supports both auto-triggered retros (after ftm-executor completes) and manual retros (via /ftm retro) to accommodate workflow preferences.
• Provides concrete improvement suggestions grounded in recorded evidence to close feedback loops.

Quick Start

Run /ftm retro after a completed plan to generate a scored retrospective report and write it to ~/.claude/ftm-retros.