GitHub Review Iteration
prisma/orm
Runs a loop on a GitHub pull request: fetch review state, triage comments into actions, implement them and resolve threads, repeating until nothing actionable is left.
A skill your agent uses when the user invokes $review-a-change or asks for a compartment repo code review; always use the delegated subagent review path by splitting review into separate parallel…
Install Review A Change
$ npx skills add compartmentdev/compartment --skill review-a-change -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install compartmentdev/compartment review-a-change --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/compartmentdev/compartment.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.codex/skills/review-a-change .claude/skills/review-a-change && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "review-a-change" agent skill from https://github.com/compartmentdev/compartment/tree/main/.codex/skills/review-a-change into .claude/skills/review-a-change/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "review-a-change", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/compartmentdev/compartment/tree/main/.codex/skills/review-a-changeType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add compartmentdev/compartment --skill review-a-change -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install compartmentdev/compartment review-a-change --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/compartmentdev/compartment.git skills-src && mkdir -p .agents/skills && cp -r skills-src/.codex/skills/review-a-change .agents/skills/review-a-change && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "review-a-change" agent skill from https://github.com/compartmentdev/compartment/tree/main/.codex/skills/review-a-change into .agents/skills/review-a-change/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "review-a-change", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add compartmentdev/compartment --skill review-a-change -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install compartmentdev/compartment review-a-change --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/compartmentdev/compartment.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/.codex/skills/review-a-change .cursor/skills/review-a-change && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "review-a-change" agent skill from https://github.com/compartmentdev/compartment/tree/main/.codex/skills/review-a-change into .cursor/skills/review-a-change/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "review-a-change", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/compartmentdev/compartment.git --path .codex/skills/review-a-change--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add compartmentdev/compartment --skill review-a-change -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install compartmentdev/compartment review-a-change --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/compartmentdev/compartment.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/.codex/skills/review-a-change .gemini/skills/review-a-change && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "review-a-change" agent skill from https://github.com/compartmentdev/compartment/tree/main/.codex/skills/review-a-change into .gemini/skills/review-a-change/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "review-a-change", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install compartmentdev/compartment review-a-changeInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add compartmentdev/compartment --skill review-a-change -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/compartmentdev/compartment.git skills-src && mkdir -p .github/skills && cp -r skills-src/.codex/skills/review-a-change .github/skills/review-a-change && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "review-a-change" agent skill from https://github.com/compartmentdev/compartment/tree/main/.codex/skills/review-a-change into .github/skills/review-a-change/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "review-a-change", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add compartmentdev/compartment --skill review-a-change -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install compartmentdev/compartment review-a-change --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/compartmentdev/compartment.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/.codex/skills/review-a-change .opencode/skills/review-a-change && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "review-a-change" agent skill from https://github.com/compartmentdev/compartment/tree/main/.codex/skills/review-a-change into .opencode/skills/review-a-change/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "review-a-change", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
Facts
- Skill name
review-a-change- GitHub stars
- 205
- Token cost
- ~1.6k tokens
- SKILL.md length
- 862 words
- Files
- 2
- Skills in repo
- 12
- Repo updated
- First seen
- Licence
- Apache-2.0
At a glance
A skill your agent uses when the user invokes $review-a-change or asks for a compartment repo code review; always use the delegated subagent review path by splitting review into separate parallel…
- Works in 12 steps: Identify the owning package, entry… → Always launch separate subagents for the… → Let review subagents inherit the user's… → …
- The user invokes $review-a-change
- SKILL.md covers Invocation notes, Source of truth, Review workflow and Review focus, plus 1 more section
- Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md
What it does
Review A Change is an agent skill from compartmentdev/compartment. Use when the user invokes $review-a-change or asks for a compartment repo code review; always use the delegated subagent review path by splitting review into separate parallel passes for contracts, architecture, runtime surface, validation, and changed slices, then report findings first with file and line references.
Its SKILL.md is about 1.6k tokens, which your agent loads only when the skill is triggered. The skill folder holds 2 other files (for example `agents/openai.yaml`).
It sits in Development, covering Subagents. The repository describes itself as: Compartment is a self-hosted deployment system for small software on infrastructure your team controls. The licence is Apache-2.0.
When your agent uses it
- The user invokes $review-a-change
- Asks for a compartment repo code review
- Always use the delegated subagent review path by splitting review into separate parallel passes for contracts
- Runtime surface
Example prompts
- “/review-a-change”
Requirements
- Docker
Workflow steps
12 steps, taken from the first numbered list in SKILL.md.
- Identify the owning package, entry boundary, contracts, runtime surface, and the single canonical path touched by the change.
- Always launch separate subagents for the independent passes below and keep the main agent as the integrator. Do not choose or name a…
- Let review subagents inherit the user's current session configuration. If the client requires explicit values, pass only the current…
- Start these required subagents in parallel. If the client only allows sequential spawn calls, start them one after another before reviewing
- Split the remaining changed surface into 1 to 3 additional slice subagents by owning package or boundary, not by arbitrary file count. Ask…
- For tiny diffs, still run the contracts, runtime-hygiene, and validation subagents plus at least one slice subagent. Do not collapse back…
- Tell each subagent to return findings only, with file and line references, and not to edit files or resolve findings.
- If the current agent is already one of the delegated review subagents, run only the assigned pass and do not re-delegate.
- Do not perform a main-agent-only fallback or do the same review dimensions serially.
- Treat any new fallback behavior, alternate selection path, or legacy compatibility check as a finding unless the diff or request makes…
- Treat any new or widened external exposure of /internal/*, internal-token routes, control-plane health probes, undocumented public host…
- Check architecture fit before commenting on local style.
What it can do on your machine
Read from SKILL.md and the folder at commit 15373e7. It shows what the files ask for, not the result of running them.
Tool permissions
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Runs code
No scripts in the folder and no shell commands in SKILL.md.
From the folder's file list and the shell code blocks in SKILL.md.
Network
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Credentials
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Context cost
Review A Change loads about 1.6k tokens when it runs. Until then it costs about 84 tokens; SKILL.md has 862 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
Safety
Auto-check passed
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.
SKILL.md
The full file from compartmentdev/compartment at commit 15373e7, republished under its Apache-2.0 licence (© compartmentdev). 862 words, ~1,578 tokens.
Download SKILL.mdSave it as
.claude/skills/review-a-change/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.- name
- review-a-change
- description
- Use when the user invokes $review-a-change or asks for a compartment repo code review; always use the delegated subagent review path by splitting review into separate parallel passes for contracts, architecture, runtime surface, validation, and changed slices, then report findings first with file and line references.
Review a Change
Use this skill for code review in the compartment repo.
Invocation notes
$review-a-changeis the explicit skill invocation./reviewmay be a client built-in command rather than this skill. Do not rely on/reviewalone to load this skill or to grant subagent permission.- This skill requires the delegated subagent path for every review. Do not run a main-agent-only fallback.
Source of truth
AGENTS.mddocs/layers/README.md- the short layer doc for each touched owning package
docs/specs/type-placement.mdwhen types move or ownership is part of the review$public-docs-maintenanceonly when public docs may be required
Review workflow
- Identify the owning package, entry boundary, contracts, runtime surface, and the single canonical path touched by the change.
- Always launch separate subagents for the independent passes below and keep the main agent as the integrator. Do not choose or name a model, reasoning effort, speed, latency, or quality mode in this skill.
- Let review subagents inherit the user's current session configuration. If the client requires explicit values, pass only the current user-selected values exposed by the client/runtime. Prefer a subagent type that preserves that configuration.
- Start these required subagents in parallel. If the client only allows sequential spawn calls, start them one after another before reviewing:
- contracts-and-architecture: package ownership, layer fit, contracts, type placement, boundary drift, API shape, hidden fallback paths, and the documented public ingress contract. This pass must run
$audit-type-ownershipeven when you expect no findings. - runtime-hygiene: exports, entrypoints, scaffolding, dead runtime surface, hidden compatibility branches, and public-surface expansion such as new host aliases or externally reachable internal paths. This pass must run
$compartment-runtime-hygieneeven when runtime changes look incidental. - docs-drift: missing or misleading public docs only when users must change behavior, understand a public contract, or make an operator decision; generated reference drift; and proposal-only language leaking into the public site. This pass must use
$public-docs-maintenanceas the policy source and must not edit docs. - validation: missing tests, wrong test depth, missing migrations, missing fixtures, and missing DB or Docker checks.
- contracts-and-architecture: package ownership, layer fit, contracts, type placement, boundary drift, API shape, hidden fallback paths, and the documented public ingress contract. This pass must run
- Split the remaining changed surface into 1 to 3 additional slice subagents by owning package or boundary, not by arbitrary file count. Ask each slice pass to look only for bugs, regressions, clean-code issues, oversized functions, and KISS drift inside its slice.
- For tiny diffs, still run the contracts, runtime-hygiene, and validation subagents plus at least one slice subagent. Do not collapse back to one slow serial review.
- Tell each subagent to return findings only, with file and line references, and not to edit files or resolve findings.
- If the current agent is already one of the delegated review subagents, run only the assigned pass and do not re-delegate.
- Do not perform a main-agent-only fallback or do the same review dimensions serially.
- Treat any new fallback behavior, alternate selection path, or legacy compatibility check as a finding unless the diff or request makes that path explicit and necessary.
- Treat any new or widened external exposure of
/internal/*, internal-token routes, control-plane health probes, undocumented public host aliases, or internal/public URL mixing as a blocking security finding, not an advisory note. - Check architecture fit before commenting on local style.
- Review the smallest correct surface, not just whether the code works.
- Integrate all passes into one deduplicated findings list ordered by severity, then do one short main-agent sweep for cross-slice interactions that the parallel passes could miss.
- If a review finds a real issue that should not be fixed in the current PR, keep it explicit in the review findings or PR discussion. Create a GitHub follow-up issue only when the user asks for issue tracking or the current PR workflow explicitly requires it.
Show full SKILL.md (244 more words)Show less
Review focus
- architecture or layer ownership drift;
- guessed, broadened, or duplicated contracts, including
id | slug | namefallback matching and public types hidden behind inference; - hidden or unagreed fallback paths, speculative compatibility branches, and duplicated legacy checks added without a stated contract;
- KISS regressions where the change introduces extra branching or duplicate logic instead of keeping one canonical path;
- production APIs widened for tests instead of using package-local seams;
- clean-code regressions such as muddled orchestration, hidden state coupling, or misleading output/debug surfaces;
- public ingress drift such as externally reachable internal URLs, internal-token routes on public hosts, control-plane health probes exposed through public ingress, internal callers using public URLs, or undocumented host aliases and wildcard expansion;
- docs drift where
AGENTS.mdrequires public docs but they are absent or misleading; - oversized or unsplit source files that hide the real change surface;
- dead runtime surface from exports, entrypoints, or scaffolding;
- missing DB migrations when schema or persisted shape changed;
- missing tests or the wrong validation depth for the scope.
Output
- Report findings first, ordered by severity, with file and line references.
- Focus on bugs, regressions, architecture drift, and missing tests or checks.
- Call out hidden fallback or legacy paths explicitly instead of treating them as harmless defensive code.
- If there are no findings, say so explicitly and mention any residual risk briefly.
- When you intentionally defer a valid issue, make the deferment explicit in the review output and note any follow-up issue only if one was actually created.
© compartmentdev, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
Files
SKILL.md and 1 other file in .codex/skills/review-a-change of compartmentdev/compartment.
- SKILL.md
- agents/openai.yaml
Open the folder on GitHubat commit 15373e7
Compare with similar skills
Review A Change next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Review A Change this skillcompartmentdev/compartment | 205 | — | ~1.6k | Automated safety check: Pass | Apache-2.0 | |
| GitHub Review Iterationprisma/orm | 48k | — | ~2.2k | Automated safety check: Pass | Apache-2.0 | |
| Cherry Studio PR ReviewCherryHQ/cherry-studio | 52k | — | ~3.9k | Automated safety check: Pass | AGPL-3.0 | |
| Jevgrepdzhng/jevgrep | 2.4k | — | ~741 | Automated safety check: Pass | MIT | |
| PR Cyclejaemk/cached | 2.1k | — | ~4.8k | Automated safety check: Notes | MIT | |
| PR Reviewjaemk/self_update | 961 | — | ~1.5k | Automated safety check: Notes | MIT |
Similar skills
- Official
Cherry Studio PR Review
CherryHQ/cherry-studio
Reviews Cherry Studio branches, pull requests, commits, files and docs against the project's own architecture, naming, API-boundary and UI rules, report-only by default.
Jevgrep
dzhng/jevgrep
A skill your agent uses for questions about how, why, or where behavior works in a repository, including questions that name a function or setting.
PR Cycle
jaemk/cached
PR review-and-update cycle — the orchestrator that takes a PR from review to resolved.
PR Review
jaemk/self_update
Targeted, read-only review of a PR or checked-out branch. An agent skill from jaemk/self_update.
Cursor Composer Task Delegate
Chachamaru127/claude-code-harness
Hands one implementation task to Cursor Composer in an isolated git worktree, then reviews its diff and cherry-picks the result into the main branch.
More from compartmentdev/compartment
All 12 skills in this repoAudit Type Ownership
compartmentdev/compartment
Audit type ownership and boundary placement in the compartment repo: misplaced or duplicated types, query/service leaks, root export surface, and phased move plans.
Open PR And Monitor
compartmentdev/compartment
Open or reuse a repo PR and monitor it with a five-minute Codex heartbeat until it is ready, terminal, or blocked.
Review Reuse And Dedup
compartmentdev/compartment
A skill your agent uses when reviewing compartment repo changes for existing helper reuse, extractable pure utilities, duplicated logic, repeated constants, hardcoded values, or repeated variable…
Compartment Smoke Test
compartmentdev/compartment
Run or delegate a browser-first smoke test of the local compartment through browser-use as a real user: boot or reuse the local stack, install or log in, deploy the repo smoke fixtures, verify CLI…
Develop A Feature
compartmentdev/compartment
Implement repo changes by starting with preflight feature review, then editing the smallest owning surface on one canonical path and running only diff-triggered checks.
Public Docs Maintenance
compartmentdev/compartment
Maintain the public Starlight docs site for material changes to shipped Compartment workflows, using generated reference where possible and concise task- and decision-focused guides elsewhere.
Related
Categories
Questions about Review A Change
What does Review A Change do?
A skill your agent uses when the user invokes $review-a-change or asks for a compartment repo code review; always use the delegated subagent review path by splitting review into separate parallel…. Review A Change is an agent skill from compartmentdev/compartment. Use when the user invokes $review-a-change or asks for a compartment repo code review; always use the delegated subagent review path by splitting review into separate parallel passes for contracts, architecture, runtime surface, validation, and changed slices, then report findings first with file and line references.
When should I use Review A Change?
Review A Change fits situations like: the user invokes $review-a-change; asks for a compartment repo code review; always use the delegated subagent review path by splitting review into separate parallel passes for contracts; runtime surface.
How do I install Review A Change in Claude Code?
Run `npx skills add compartmentdev/compartment --skill review-a-change -a claude-code`. Or copy the skill folder (.codex/skills/review-a-change in compartmentdev/compartment) into .claude/skills/review-a-change in your project. Claude Code loads it when a task matches its description.
How do I install Review A Change in Codex?
Run `npx skills add compartmentdev/compartment --skill review-a-change -a codex`. Or copy the skill folder (.codex/skills/review-a-change in compartmentdev/compartment) into .agents/skills/review-a-change in your project. Codex loads it when a task matches its description.
Can I use Review A Change in Cursor, Gemini CLI or GitHub Copilot?
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add compartmentdev/compartment --skill review-a-change -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/review-a-change, .gemini/skills/review-a-change, .github/skills/review-a-change and .opencode/skills/review-a-change in your project.
What does Review A Change need to run?
SKILL.md names no scripts, command-line tools or credentials: Review A Change is instructions for the agent only. Our summary lists: Docker.
Does Review A Change access the network?
SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.
Is Review A Change safe to install?
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.
What licence does Review A Change use?
Review A Change is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
How many tokens does Review A Change use?
About 1.6k tokens (SKILL.md is roughly 6.3k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.
What are the alternatives to Review A Change?
Skills that share tags, products or a category with Review A Change: GitHub Review Iteration (prisma/orm, 48k stars), Cherry Studio PR Review (CherryHQ/cherry-studio, 52k stars), Jevgrep (dzhng/jevgrep, 2.4k stars) and PR Cycle (jaemk/cached, 2.1k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Who maintains Review A Change?
compartmentdev (a GitHub organization) maintains it in compartmentdev/compartment, which has 205 GitHub stars. The repository holds 12 skills in this directory. The repository was last updated on August 20, 2026.
Source: compartmentdev/compartment on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.