agentleFS
matthewmturnermatthewmturner/skills/summarize-repo-changes/SKILL.md

summarize-repo-changes

Skill0 starsChanged 4 months agoMITView on GitHub

Analyze commits to a repository's default branch (main or master) within a given time interval and produce a structured summary of changes. Focuses on new or deprecated features, API changes, configuration changes, and Kubernetes/Helm resource changes. Use when the user wants to understand what changed in a repo over a recent period.

  • Reads credentials

The file

Show its rules and text

We split this skill into the rules it gives an agent

13 rules. About 0 are common, 8 are distinctive, and 5 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    If gh repo view fails with a 404 or permission error, report the error and ask the user to authenticate or provide a local path.

    line 54
  2. Distinctive:

    If the interval phrase cannot be mapped, ask the user to rephrase.

    line 90
  3. Distinctive:

    High-volume handling: If there are more than 50 commits, filter out likely trivial changes by excluding subjects matching:

    line 101
  4. Distinctive:

    Check if the repository contains Kubernetes manifests or Helm charts:

    line 134
  5. Distinctive:

    If any patterns match, set a flag to include a dedicated Kubernetes / Helm Changes section in the summary.

    line 149
  6. Very specific:

    Commit prefixes: feat:, feat(

    line 156
  7. Distinctive:

    If the repo follows Conventional Commits:

    line 184
  8. Very specific:

    Private repos / no auth: If gh auth status fails or gh repo view returns 404, tell the user to authenticate or provide a local path to the …

    line 247
  9. Distinctive:

    Interval parse failure: Ask the user to rephrase the interval.

    line 252
  10. Distinctive:

    Shallow clone: If git log --since fails due to shallow history, re-fetch with greater depth or do git fetch --unshallow.

    line 253
  11. Very specific:

    When Chart.yaml version changes, always call out the old version and new version explicitly.

    line 265
  12. Very specific:

    If the repo uses kustomize (look for kustomization.yaml), include kustomize overlay/base changes in the Kubernetes section.

    line 267
  13. Distinctive:

    Keep the summary scannable — use bullet points and bold callouts for breaking changes.

    line 268

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Summarize Repo Changes

Analyze all commits on a repository's default branch within a specified time interval and produce a structured markdown summary of meaningful changes.

Prerequisites

  • git installed and available
  • gh CLI installed and authenticated (gh auth status) — required for GitHub repos
  • Access to the target repository (clone permissions for GitHub repos, filesystem access for local repos)

Usage

The user provides:

ParameterRequiredDescription
repoYesowner/repo for GitHub repos, or a local filesystem path to a git repo
intervalYesTime window string (e.g., "8 hours", "1 week", "3 days", "since Monday")
branchNoOverride the default branch detection (e.g., "develop", "trunk")

Workflow

Step 1: Resolve Repository

Determine whether the input is a GitHub repo or a local path.

GitHub repo (matches owner/repo pattern):

# Verify access and get repo metadata
gh repo view owner/repo --json defaultBranchRef,cloneUrl,description

# Clone to a temp directory if not already available
REPO_DIR=$(mktemp -d)
git clone --depth 50 --branch BRANCH CLONE_URL "$REPO_DIR"
cd "$REPO_DIR"
# Fetch full history for the branch to analyze all commits
git fetch origin BRANCH --depth=1000 2>/dev/null || true

Local repo (filesystem path):

# Validate it is a git repo
git -C /path/to/repo rev-parse --git-dir
cd /path/to/repo

If gh repo view fails with a 404 or permission error, report the error and ask the user to authenticate or provide a local path.

Step 2: Determine Default Branch

# GitHub: use the default branch from gh repo view
# Extract defaultBranchRef.name from JSON output

# Local: try upstream HEAD, then fall back to main, then master
git symbolic-ref refs/remotes/origin/HEAD 2>/dev/null | sed 's|refs/remotes/origin/||'
git rev-parse --verify main  >/dev/null 2>&1 && echo "main" || echo "master"

Allow a user-provided branch parameter to override detection.

Step 3: Compute Time Range

Map the natural language interval to a date -d expression:

Interval Phrasedate -d Expression
"8 hours" / "8h"8 hours ago
"1 day" / "24 hours"1 day ago
"3 days" / "last 3 days"3 days ago
"1 week" / "7 days"7 days ago
"2 weeks"14 days ago
"1 month" / "30 days"30 days ago
"since Monday"this Monday
"since yesterday"yesterday
# Compute the since timestamp
SINCE=$(date -d "EXPRESSION" --iso-8601=seconds)
NOW=$(date --iso-8601=seconds)

If the interval phrase cannot be mapped, ask the user to rephrase.

Step 4: List Commits in Interval

# Get commit list: hash, date, subject
git log --oneline --since="$SINCE" --until="$NOW" --format="%H %ai %s" BRANCH

Capture all commit hashes, dates, and subject lines for analysis.

High-volume handling: If there are more than 50 commits, filter out likely trivial changes by excluding subjects matching:

# Patterns indicating trivial changes
grep -iEv '^\w+ +(typo|format|whitespace|lint|chore|bump dep|update lock|regen|revert of revert|merge branch)' 

Still count all commits in the report but focus detailed analysis on the remaining set.

Step 5: Fetch Commit Details

For each meaningful commit, get the file-level change summary:

git show --stat --format="%H %an %ai %s" COMMIT

For the top ~20 most significant commits (judged by title keywords, number of files changed, or file paths touched), get the full diff:

git diff COMMIT~1 COMMIT -- . ':(exclude)node_modules/' ':(exclude)vendor/' ':(exclude)dist/' ':(exclude)build/'

For GitHub repos with merge commits, enrich with PR details:

gh pr view PR_NUMBER --repo owner/repo --json title,body,labels,mergedAt,reviewers,files

Extract the PR body and labels for richer context (breaking-change labels, feature labels, etc.).

Step 6: Detect Kubernetes / Helm Usage

Check if the repository contains Kubernetes manifests or Helm charts:

# Helm detection
ls charts/ Chart.yaml values.yaml 2>/dev/null
find . -name "Chart.yaml" -o -name "values.yaml" -o -name "values-*.yaml" 2>/dev/null | head -20

# Kubernetes manifest detection (directories)
ls k8s/ deploy/ deployments/ manifests/ overlays/ base/ 2>/dev/null

# Kubernetes manifest detection (file content)
grep -rl "^apiVersion:.*" --include="*.yaml" --include="*.yml" . 2>/dev/null | \
  xargs grep -l "^kind:.*" 2>/dev/null | head -30

If any patterns match, set a flag to include a dedicated Kubernetes / Helm Changes section in the summary.

Step 7: Categorize Changes

Apply these heuristics to classify each commit:

New Features

  • Commit prefixes: feat:, feat(
  • Title keywords: add, new, introduce, implement, create
  • New files added (check git diff --diff-filter=A)
  • New directories or modules created

Deprecated / Removed Features

  • Title keywords: deprecat, remove, drop, retire, eliminate, phase out
  • Files deleted (check git diff --diff-filter=D)
  • Removed exports, modules, or public APIs

API Changes

  • Files changed: routes, handlers, controllers, api/, grpc/, protobuf/, proto/, swagger*, openapi*, *.proto
  • Title keywords: API, endpoint, endpoint, breaking, BC break, backward incompatibl
  • Body/footer markers: BREAKING CHANGE:, BREAKING-CHANGE:

Configuration Changes

  • Files changed: *.env*, config.*, settings.*, .env.*, application.*.yml, *.conf, Dockerfile, docker-compose*, *.toml, *.ini
  • New or changed ENV / ARG instructions in Dockerfiles
  • New environment variable references in code (search diff for process.env., os.environ, os.Getenv, os.Getenv, config., cfg.)

Kubernetes / Helm Changes (only if K8s/Helm detected in Step 6)

  • Files changed: *.yaml with apiVersion:/kind:, Chart.yaml, values.yaml, values-*.yaml, *.tpl in templates/
  • Chart version bumps (compare old/new version in Chart.yaml)
  • Changes to: Deployments, Services, Ingress, CRDs, RBAC (ClusterRole/Role/Binding), ConfigMaps, Secrets, HPA, NetworkPolicies
  • Template logic changes in Helm *.tpl files
  • Breaking manifest changes (e.g., removed resources, changed API versions, renamed selectors)

Conventional Commits Shortcuts If the repo follows Conventional Commits:

  • feat: → New Features
  • fix: → Bug fixes (include in notable commits)
  • BREAKING CHANGE: → call out prominently under API Changes
  • chore:, docs:, style:, test: → typically skip unless file paths indicate config or infra changes
  • perf:, refactor: → include if they affect public interfaces

Step 8: Generate Summary

Produce the following markdown output:

# Repository Changes: REPO

**Period:** SINCE → NOW (N commits analyzed)
**Branch:** BRANCH

## Summary

1-2 sentence high-level overview of what happened in this period.

## New Features

- Description of each new feature, linked to the commit or PR
- Note the scope and files affected

## Deprecated / Removed

- Description of deprecations or removals
- Note any migration guidance mentioned in commits

## API Changes

- Breaking changes (call out with **BREAKING** prefix)
- New endpoints or API surfaces
- Parameter or behavior changes

## Configuration Changes

- New or changed environment variables
- Config file changes and their impact
- Docker/container configuration changes

## Kubernetes / Helm Changes

_(This section only appears if K8s/Helm resources were detected.)_

- Chart version bumps (old → new)
- Manifest changes by resource type (Deployments, Services, CRDs, etc.)
- values.yaml default value changes
- Breaking manifest changes (**BREAKING** callout)
- Ingress or networking changes

## Notable Commits

| Commit | Author | Date | Message | Files |
|--------|--------|------|---------|-------|
| `abc1234` | @author | 2026-05-27 | feat: add new auth provider | 5 |
| `def5678` | @author | 2026-05-27 | fix: resolve race condition | 2 |

Edge Cases

  • Private repos / no auth: If gh auth status fails or gh repo view returns 404, tell the user to authenticate or provide a local path to the repo.
  • No commits in interval: Output: "No commits found on BRANCH between SINCE and NOW."
  • Very large diffs (100+ commits): Summarize at stat level for all commits; sample full diffs from at most 20 of the most significant commits. Note the sampling in the summary.
  • Local repos with no remote: Skip all gh commands; use git only for analysis.
  • Merge commits: Include the merge commit message (which typically contains the PR title). When gh is available, also fetch the PR body for context.
  • Interval parse failure: Ask the user to rephrase the interval.
  • Shallow clone: If git log --since fails due to shallow history, re-fetch with greater depth or do git fetch --unshallow.

Tips

  • For repos with a CHANGELOG.md or CHANGELOG.md, check if recent entries align with the interval and reference them as additional context:

    # Check if changelog exists and has recent entries
    head -100 CHANGELOG.md 2>/dev/null
  • Cross-reference merge commits with gh pr view to pick up labels like breaking-change, enhancement, bug for accurate categorization.
  • When Chart.yaml version changes, always call out the old version and new version explicitly.
  • For Helm template changes (*.tpl files), note if conditional logic ({{ if }}, {{ range }}) changed since this can silently alter deployed resources.
  • If the repo uses kustomize (look for kustomization.yaml), include kustomize overlay/base changes in the Kubernetes section.
  • Keep the summary scannable — use bullet points and bold callouts for breaking changes.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.