agentsclimarketplace

Git period report

Skill forchain/skills/git-period-report

Personal Codex and agent skills collection

Install
npx -y skills add forchain/skills --skill git-period-report

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • no licenseNo license file was found in the repository. Code published without one is not open source by default, so using it at work is a question for whoever answers licensing questions where you are.
  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Generate git-based period reports from repository history: weekly reports by default, monthly reports, custom date-range summaries, commit-history reports, and squashed-diff/final-state reports. Use when the user asks for a weekly report, monthly report, work report, changelog-like status report, or asks to summarize commits or changes over a time window.

SKILL.md

3.2 KB, 646 tokens by cl100k_base, as published. Nobody here has run it

Git Period Report

Overview

Generate a weekly, monthly, or custom period report from a Git repository. Treat the squashed diff between the period's start baseline and end snapshot as the source of truth; use commits as trace evidence.

Default period: the last completed ISO week, Monday through Sunday.

Workflow

  1. Resolve the period

    • Use --period weekly when the user does not specify a period.
    • Use --period monthly for a completed previous-month report.
    • Use --since YYYY-MM-DD --until YYYY-MM-DD for a custom range.
    • Completion criterion: the report has exact inclusive start and end dates.

  2. Extract git context

    • Run: 
      python3 <skill-dir>/scripts/git_period_context.py --repo . --period weekly --output /tmp/git-period-report.json
    • For monthly: 
      python3 <skill-dir>/scripts/git_period_context.py --repo . --period monthly --output /tmp/git-period-report.json
    • For custom: 
      python3 <skill-dir>/scripts/git_period_context.py --repo . --since 2026-06-01 --until 2026-06-30 --output /tmp/git-period-report.json
    • Completion criterion: the JSON contains base_commit, end_commit, shortstat, changed files, and period commits.

  3. Read the final-state diff

    • Use the JSON as a map, then inspect important files with: 
      git -C <repo> diff <base_commit> <end_commit> -- <path>
git -C <repo> show <end_commit>:<path>
    • Give priority to files with high churn, user-facing paths, runtime-critical paths, migrations, tests, docs, and configuration.
    • Completion criterion: every major theme in the report is backed by final diff evidence, not only a commit title.

  4. Write the report

    • Read references/report-writing.md before drafting.
    • Organize by final capabilities and user impact. Do not list every commit unless the user specifically asks for a commit ledger.
    • Include the exact range, baseline commit, end commit, and diff scale.
    • Completion criterion: the report explains what changed, why it mattered, and what residual risks or follow-ups remain.

  5. Verify saved output

    • If writing into a repository, run git diff --check.
    • If the repository has a docs build or markdown lint command, run the relevant lightweight check.
    • Completion criterion: report file is saved where requested, and verification results are stated.

Notes

  • For weekly reports, prefer the last completed week over the current partial week.
  • For monthly reports, prefer the last completed calendar month over the current partial month.
  • If the user explicitly asks for current week/month-to-date, honor that override and label the report as partial.
  • If no commit exists before the start boundary or before the end boundary, stop and report the missing boundary.

What ships with it: 3 files

9.8 KB alongside SKILL.md, 1 of them executable

agents/

references/

scripts/

Keep looking

Skills are one crate of 327,132. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.