light-load skill
✓ VerifiedImproves pull request reviewability by cleaning commit history, sharpening descriptions, and adding reviewer guidance without changing code behavior.
- GitHub stars
- 113
- Files
- 1
- Last commit
- Nov 1, 2026
- License
- MIT
- Allowed tools
- —
What does the light-load skill do?
light-load improves pull request reviewability by cleaning commit history, sharpening descriptions, and adding reviewer guidance without changing code behavior.
Summary generated from the skill's SKILL.md.
About the light-load skill
Preparing pull requests for easier inspection is what light-load accomplishes for developers. It functions by analyzing commit history, diff sizes, and changed paths to identify noisy or disorganized contributions. An agent should load this skill when a user requests to tidy a PR, clean up commits, or annotate a diff to make the intent clearer for reviewers.
The tool focuses on improving reviewability without altering the actual code behavior. It can propose plans to rewrite history into logical groupings, such as separating schema changes from UI behavior, or it can enhance descriptions using tea-time and yip-yip formats. It relies on git commands and the GitHub CLI to inspect PR details and verify that tree hashes remain identical after any history rewrites. If a pull request is too massive to fix with simple notes, the skill recommends splitting the work into a stack instead of attempting to polish a problematic diff.
What an agent can do with it
- 1Clean up commit historyOrganize messy commits into a logical dependency order like schema, core logic, and tests.
- 2Improve PR descriptionsEnhance the pull request summary to better explain the diff and call out risky behavior.
- 3Verify history rewritesEnsure that rewriting git history does not unintentionally change the underlying code content.
- 4Manage oversized pull requestsIdentify when a pull request is too large to be effectively reviewed and suggest splitting it.
Written from the skill's SKILL.md. Every use case is checked against a quote from it.
How to install the light-load skill
Each command copies the skill folder into the client's skills directory. Restart the client or start a new session to load it.Install the light-load skill in Plugin
/plugin marketplace add tmcfarlane/team-avatar /plugin install team-avatar@team-avatar
Install the light-load skill in Claude Code
~/.claude/skills/light-load/SKILL.md
$ npx degit tmcfarlane/team-avatar/plugins/team-avatar/skills/light-load ~/.claude/skills/light-loadInstall the light-load skill in Claude Code (project)
.claude/skills/light-load/SKILL.md
$ npx degit tmcfarlane/team-avatar/plugins/team-avatar/skills/light-load .claude/skills/light-loadInstall the light-load skill in Codex
~/.agents/skills/light-load/SKILL.md
$ npx degit tmcfarlane/team-avatar/plugins/team-avatar/skills/light-load ~/.agents/skills/light-loadInstall the light-load skill in Gemini CLI
~/.gemini/skills/light-load/SKILL.md
$ npx degit tmcfarlane/team-avatar/plugins/team-avatar/skills/light-load ~/.gemini/skills/light-loadInstall the light-load skill in Claude.ai
npx degit tmcfarlane/team-avatar/plugins/team-avatar/skills/light-load light-load && zip -r light-load.zip light-load # claude.ai → Customize → Skills → + → Create skill → Upload a skill → light-load.zip
Automated checks
Ran Nov 14, 2026Full report →Passed our automated checks. That is not a security audit: read the scripts before you let an agent run them. How checks work
- Source repository linked and reachabletmcfarlane/team-avatar/plugins/team-avatar/skills/light-load is public and reachablegithub.com/tmcfarlane/team-avatar/tree/HEAD/plugins/team-ava ↗Pass
- Open source license (OSI)MITspdx.org/licenses/MIT.html ↗Pass
- Commit or release in the last yearLast commit 2026-11-01github.com/tmcfarlane/team-avatar/commits ↗Pass
- No download-and-run or credential theft in scriptsNo scripts in the skill folderPass
- No hidden instructions in SKILL.mdNo override phrases, safeguard switches, credential requests or hidden charactersgithub.com/tmcfarlane/team-avatar/blob/HEAD/plugins/team-ava ↗Pass
- Pre-approved tools are scopedNo pre-approved tools; the agent asks before using onegithub.com/tmcfarlane/team-avatar/blob/HEAD/plugins/team-ava ↗Pass
- SKILL.md follows the Agent Skills formatValid name and description; says when to use it; 65 linesgithub.com/tmcfarlane/team-avatar/blob/HEAD/plugins/team-ava ↗Pass
SKILL.md
View on GitHubAppa's light load
Appa can carry the whole team, but he flies farthest when the saddle holds only what the trip needs, packed so anyone can find what they are looking for. A PR is the same. Prepare it so a reviewer quickly understands the intent, the important files, and the risk. The default goal is reviewability with no behavior change.
Steps
- Resolve the target PR. Use the URL the user gave, or the PR for the current branch.
- Inspect it. Read the commits, the diff size, the changed paths, the generated files, and the PR description.
- Name the reviewability problems. Look for noisy commits, a stale description, unrelated changes, mechanical and logic changes mixed together, missing tests, or no clear place for the reviewer to start.
- Propose a plan before rewriting history or force-pushing. Wait for the user's agreement on any history rewrite.
- Apply the safe improvements. Then verify that the tree or diff still matches the intended code.
History cleanup
Rewrite history only when the user asks for it or agrees to the plan. Before you rewrite, record the original tree:
gh pr view <PR> --json title,headRefName,baseRefName,state,commits
git fetch origin <headRefName> <baseRefName>
ORIGINAL_TREE=$(git rev-parse origin/<headRefName>^{tree})
Good commit groupings usually follow dependency order:
- Schema, storage, or generated API definitions.
- Core logic.
- Wiring and integration.
- UI or surface behavior.
- Tests.
After you rewrite, verify that the content is identical:
echo "Original tree: $ORIGINAL_TREE"
echo "Current tree: $(git rev-parse HEAD^{tree})"
git diff origin/<headRefName> --stat
Do not push if the tree changed unintentionally. An empty git diff and matching tree hashes are the proof.
Reviewer guidance
When code behavior must stay untouched, improve the PR description and add review notes instead. Write them with tea-time, in the description shape from yip-yip.
- Add a short summary at the top that matches the actual diff.
- Separate the core files from generated or mechanical files.
- Call out risky behavior changes, migration order, the rollout plan, and test coverage.
- Link issue trackers, dashboards, or design docs when they explain intent.
Rules
- Never hide a meaningful behavior change inside "cleanup".
- Do not bypass hooks (
--no-verify) unless the user explicitly asks. - Never force-push a branch others may have pulled without the user's word.
- If the PR is too large to make reviewable with notes, recommend splitting it into a stack (see
yip-yip) instead of polishing around the problem.
Reply
The problems you found, what you changed, the before and after tree hashes when history moved, and anything you recommend but did not do.
Files in this skill
1light-load skill questions
How do I install the light-load skill in Claude Code?
Run npx degit tmcfarlane/team-avatar/plugins/team-avatar/skills/light-load ~/.claude/skills/light-load. Claude Code loads skills from ~/.claude/skills (all projects) or .claude/skills (one project) and uses a skill when its description matches the task.
Which clients can use the light-load skill?
Any client that reads the Agent Skills format: Claude Code, Codex, Gemini CLI and claude.ai (upload as a ZIP). Scripts in the skill run only where the client allows code execution.
Does this tool change the actual code behavior?
No, the primary goal is to improve reviewability without changing the code behavior.
What should I do if a pull request is too large to be made reviewable with notes?
The tool will recommend splitting the pull request into a stack rather than attempting to polish it.