retro

Generates engineering retrospective reports from local git history analytics.

Updated Jul 9, 2026
One-click install
npx skills add https://github.com/ByronWilliamsCPA/plugin --skill retro-byronwilliamscpa
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: retro
Source: https://github.com/ByronWilliamsCPA/plugin/tree/main/plugins/wff-code/skills/retro
Command: npx skills add https://github.com/ByronWilliamsCPA/plugin --skill retro-byronwilliamscpa

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Teams closing a sprint or release often lack data-backed evidence for retrospectives, relying on anecdotes instead of measurable signals like commit cadence, churn hotspots, and PR cycle time. ## Core Features & Use Cases - Git History Analytics: Computes commit cadence, per-author contribution, churn hotspots, and Conventional Commit type distribution using only local git commands. - Trend Deltas: Compares the current window against an equal-length prior window to surface directional shifts in velocity and commit-type mix. - PR Cycle Time: Derives approximate branch lifetime from merge-commit history, or explicitly flags it as not derivable on squash-only histories. - Use Case: At the end of a two-week sprint, run a retro over the last 14 days to confirm that a file like session.py is a recurring churn hotspot and that the commit mix shifted toward fix: after a release. ## Quick Start Ask the assistant to run a retro over the last 14 days of git history and produce an engineering retrospective report.

Frequently Asked Questions about retro

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

FAQPage Schema
How do I generate an engineering retrospective from git history?▼

Run git log commands over a confirmed date window to compute commit cadence, per-author counts, churn hotspots, and commit-type distribution, then compare against an equal-length prior window. The retro skill assembles these into a structured report.

How to find churn hotspot files in a git repository?▼

Use git log with --name-only over the window, count occurrences per file, and sort descending to find the most-touched files. Pair with --numstat to see added and removed line totals for magnitude.

Can I measure PR cycle time from git history alone?▼

Cycle time is derivable only from merge-commit history, by diffing a branch's first commit against its merge date. On squash-only histories it cannot be derived from git, and the report should say so rather than invent a number.

Does this retro analysis require network access or external services?▼

No, everything comes from the local git repository already on disk. No remote calls, APIs, or external services are involved in computing the metrics.

When should I not use per-author commit counts?▼

Per-author counts should never be used as a performance ranking or productivity score. They describe where work landed, which is shaped by tasking and review load, and treating them as targets invites Goodhart's law gaming.