Score ftm-executor plan performance across five dimensions and write retro reports.
npx skills add https://github.com/kkudumu/feed-the-machine --skill ftm-retro
Please help me install this Agent Skill. Skill: ftm-retro Source: https://github.com/kkudumu/feed-the-machine/tree/main/ftm-retro Command: npx skills add https://github.com/kkudumu/feed-the-machine --skill ftm-retro
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.
Run /ftm retro after a completed plan to generate a scored retrospective report and write it to ~/.claude/ftm-retros.
High-intent search queries and answers about installing and using this skill.
Automated retrospective analysis evaluates post-execution performance by scoring wave parallelism, audit outcomes, and retry counts to generate evidence-backed insights. The system writes structured retro reports to ~/.claude/ftm-retros and updates a pattern library to guide future improvements.
Post-execution pattern analysis identifies recurring execution bottlenecks by evaluating audit outcomes and codex gate results. It captures lessons in a cumulative pattern library to provide evidence-backed improvement suggestions, making future plan executions smarter and more reliable.
Run a manual retrospective on a completed plan by executing the /ftm retro command. This triggers the scoring mechanism across five dimensions, writes a formal retro report to the local state directory, and updates patterns.json to capture recurring lessons.
Auto-triggered retrospectives support continuous quality improvement by running automatically after plan execution completes. This accommodates workflow preferences by supporting both auto-triggered runs and manual execution via the /ftm retro command without requiring manual intervention.
An automated execution retrospective scores performance across five dimensions: wave parallelism, audit pass rate, codex gate results, retry count, and overall execution smoothness. Each score includes explicit citations to specific recorded data to guide future improvements.
The retrospective report requires frontmatter-defined metadata to function correctly. The skill enforces this metadata structure during auto-triggered and manual retro runs to ensure accurate scoring and proper persistence of reports to the local state directory.
boxpReview Git-managed code changes using the codex CLI.
clovernguyen1010-ship-itCoordinate structured Agile epic retrospectives to extract lessons and capture action items.
galaxyprojectRoute Nextflow conversion tasks to the appropriate Galaxy sub-skill.
semicolon-devteamDetect and prioritize technical debt in TypeScript codebases.
