agentleFS
EpicGamesEpicGames/lore/.claude/skills/review-change/SKILL.md

review-change

Skill9.1k starsChanged 10 days agoMITView on GitHub

Must be used before submitting any change to Lore, and to review another author's change. Reviews the change adversarially, in a fresh session, against code-review.md and the standards it names. On your own branch it fixes the findings that have one obvious fix and reports the rest; on another author's change it reports the findings by severity.

29 other projects use a version of this.

Adopted by

Steady: 0 new projects took it in the last 4 weeks, none the 4 weeks before.thin data

Week of Sep 20: 0 new projectsWeek of Sep 27: 0 new projectsWeek of Oct 4: 0 new projectsWeek of Oct 11: 0 new projectsWeek of Oct 18: 0 new projectsWeek of Oct 25: 0 new projectsWeek of Nov 1: 0 new projectsWeek of Nov 8: 0 new projects
Sep 2026Oct 2026Nov 2026
New projects taking a version of this file each week, over the 8 weeks our latest run read: only 0 in all, so this is a hint, not a trend.

Most recently taken up by

Who's talking about EpicGames/lore

The biggest 5 of 39 mentions linking to the repository that our crawl found. Each line is their own headline.

See all 39

Talked about on GitHub, Bluesky and Hacker News (39 times).

On GitHub · 26

21 more on GitHub

On Bluesky · 12

7 more on Bluesky

On Hacker News · 1

Used with

Projects with it also use plan-change, implement-change and 7 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • BanKnightSkill

    plan-change

    将已完成 spec/design 的 change 拆成可执行 plan 与 tasks。用户要把设计落成执行计划、任务清单或实现步骤时使用。

    Also in 7 projects that use this

  • Codagent-AISkill

    implement-change

    Implements an approved change in one continuous session, with tiered test runs. Use when asked to implement a change from its approved definition, or to invoke codagent:implement-change.

    Also in 7 projects that use this

  • MaheshAwasareSkill

    write-runbook

    Generate an operational runbook for a service from its code, config, and dependencies — what to do when it breaks, where to look, how to roll back, who to escalate to. Use proactively before a service goes to production, after a major incident, or when handing off operational ownership.

    Also in 3 projects that use this

  • testified-ossSkill

    incident-triage

    Triage production incidents with severity assessment, root cause hypothesis, impact analysis, and escalation recommendations. Use when analyzing outages, errors, or production issues.

    Also in 4 projects that use this

  • srclightSkill

    refactor-safely

    This skill should be used when the user asks to "refactor", "rename", "move", "change a function signature", "extract a method", "delete unused code", "clean up", or any modification that could break callers or dependents. Provides a pre-change safety checklist using srclight's graph tools to assess blast radius before touching code.

    Also in 3 projects that use this

  • Google CloudSkill

    open-pr

    Commits local changes to a new Git branch and opens a GitHub pull request through the configured GitHub MCP server. Use when proposing code changes in a new pull request. Don't use for local-only commits without a remote GitHub repository.

    Also in 4 projects that use this

The file

We split this skill into the rules it gives an agent

3 rules. 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • 0

    Common

    Most of the 25 edited versions we read keep these.

  • 2

    Distinctive

    This author's own choices. Worth a look.

  • 1

    Very specific

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

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

    Read code-review.md and every standard its Guidance names, in full and in its own session.

    line 35
  2. Distinctive:

    Read each changed function and type in full with its doc comments, each changed doc page in full, and the code the change calls and is …

    line 36
  3. Distinctive:

    Check each finding against the code yourself.

    line 45

“Common” is measured: kept by at least 6 in 10 of the 25 edited versions of this file we could read. “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

Review change

Review a change to Lore as docs/developing/code-standards/code-review.md describes. The rules live in that file and in the standards its Guidance names. This skill is the procedure for applying them, and restates only what a step needs.

Arguments

  • No arguments: review the current branch against its parent branch.
  • <branch>: review that branch, another author's change.

Findings from an earlier review, and comments already on the review, are input to verify, not conclusions.

Step 1: Establish the change

Find the branch, its base, and the change's description. For your own branch, lore branch info names the parent branch, lore branch diff <parent> lists the committed changes, and lore status lists the changes not yet committed. Review both.

Check out another author's branch in an instance of its own over the shared store, with lore clone <remote-url> <path> --branch <branch> --use-shared-store. Don't switch your own working tree to it. In that instance, lore branch diff <parent> lists the changes and names the base revision they're computed from. Compare files against that base, checked out with lore clone <remote-url> <path> --revision <base> --use-shared-store, not against the parent's latest revision, which also holds the parent's newer changes.

In a git clone, git merge-base, git diff --name-only <base> and git status give the same lists, and git worktree add checks out another author's branch.

Record the revision you review, and whether the working tree holds changes beyond it. Every finding refers to it.

Step 2: Review in a fresh session

You must spawn a subagent to review: the context that wrote the change carries its reasoning, and a review from it confirms what the author meant rather than what the code does. Give it the instance path, the branch, its base, the revision, the change's description, and the earlier findings and comments. Don't give it the reasoning behind the change.

Tell the subagent to:

  1. Load the lore-development skill, and skip its review step: the subagent is that review.
  2. Read code-review.md and every standard its Guidance names, in full and in its own session. Of system-design.md and the ADRs, read the sections and decisions for the area the change touches.
  3. Read each changed function and type in full with its doc comments, each changed doc page in full, and the code the change calls and is called from. Read from the code, not from memory.
  4. Review the change as code-review.md describes. Performance and allocations come first.
  5. Return each finding as Reporting findings in code-review.md describes, and edit no file.
  6. List the standards it read at the top of its findings.

A change that adds or edits a proposal in docs/proposals/ is also reviewed with the review-lep skill, in a second subagent given the same input, spawned beside the first. Report its findings in its own Required, Recommended and Optional buckets, after the findings by severity.

Step 3: Verify the findings

Check each finding against the code yourself. Drop a finding the code doesn't support, and say why. Don't add a finding you haven't checked.

Step 4: Match tracked work

When context/team-workspace.md exists, search the work items it describes for open and deferred items that cover each finding. Refer to a covered finding by its work item rather than fixing or filing it again. A finding a work item defers stays deferred unless the user decides otherwise. Without that file, or without access to the work items, skip this step and say so.

Step 5: Act on the findings

Your own branch

  1. Fix each finding that has one obvious fix. A defect's fix must start with a test that fails without it, as testing.md describes in What a change tests.
  2. Build and run the tests the fixes touch, as the lore-development skill describes.
  3. Record the state of the working tree again, and review again from step 2 in a new subagent, with the fixed findings as input. Repeat until a review reports nothing new, for at most three rounds, then report what remains.
  4. Report the remaining findings to the user by severity: those with several reasonable fixes, design choices and trade-offs. Give the options for each, and recommend one.
  5. File each finding left for later as a work item, where step 4 found work items to search.

Another author's change

  1. Report the findings in the conversation by severity. Don't change the author's code.
  2. Post the findings to the review only when the user asks. Post one comment per severity, and name the revision reviewed at the top of each.
  3. Refer to a point an existing comment already makes rather than repeating it.

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.