Review and advance the next easy-to-merge pull request in this repository, then continue to the next simplest PR.

MITAuto-check passedDevelopment

Install Next PR

skills CLI
$ npx skills add georgewfraser/java-language-server --skill next-pr -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install georgewfraser/java-language-server next-pr --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/georgewfraser/java-language-server.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.agents/skills/next-pr .claude/skills/next-pr && rm -rf skills-src

Use ~/.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/

Facts

Skill name
next-pr
GitHub stars
814
Token cost
~2.2k tokens
SKILL.md length
1,388 words
Files
2
Skills in repo
1
Repo updated
First seen
Licence
MIT

At a glance

Review and advance the next easy-to-merge pull request in this repository, then continue to the next simplest PR.

  • The user wants Codex to list open PRs
  • SKILL.md covers GitHub Access, Checkout Rules, Workflow and Validation Rules, plus 1 more section
  • Calls gh and git
  • Pick the most trivial low-risk PR

What it does

Next PR is an agent skill from georgewfraser/java-language-server. Review and advance the next easy-to-merge pull request in this repository, then continue to the next simplest PR. Use when the user wants Codex to list open PRs, pick the most trivial low-risk PR, establish or reuse a clean baseline on the default branch, collapse the rebased PR into a single commit on top of that branch, rerun tests and benchmarks when needed, summarize findings for approval, and then either merge, close, or request changes / clarification with the user's approval.

Its SKILL.md is about 2.2k 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 Pull requests. The repository describes itself as: Java language server using the Java compiler API. The licence is MIT.

When your agent uses it

  • The user wants Codex to list open PRs
  • Pick the most trivial low-risk PR
  • Reuse a clean baseline on the default branch
  • Collapse the rebased PR into a single commit on top of that branch

Example prompts

  • “/next-pr”

What it can do on your machine

Read from SKILL.md and the folder at commit 58daaa2. 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

    Shell commands in SKILL.md call:

    • gh
    • git

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md. Its commands use gh and git, which can reach the network depending on how they are called.

    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

Next PR loads about 2.2k tokens when it runs. Until then it costs about 124 tokens; SKILL.md has 1,388 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~124
When it runs · the whole SKILL.md, loaded when a task matches
~2.2k

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 georgewfraser/java-language-server at commit 58daaa2, republished under its MIT licence (© georgewfraser). 1,388 words, ~2,228 tokens.

Download SKILL.mdSave it as .claude/skills/next-pr/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.
name
next-pr
description
Review and advance the next easy-to-merge pull request in this repository, then continue to the next simplest PR. Use when the user wants Codex to list open PRs, pick the most trivial low-risk PR, establish or reuse a clean baseline on the default branch, collapse the rebased PR into a single commit on top of that branch, rerun tests and benchmarks when needed, summarize findings for approval, and then either merge, close, or request changes / clarification with the user's approval.

Next PR

Follow this workflow when the user wants to catch up on pull requests by taking the next easiest one first and then continuing through the queue from easiest to hardest.

GitHub Access

Use the gh CLI for GitHub interactions in this workflow.

  • Use gh pr list to enumerate open PRs.
  • Use gh pr view to inspect candidate PRs and changed files.
  • Use gh pr checkout or gh api for PR-specific fetch and checkout operations.

Only fall back to the web or raw API calls if gh is unavailable or not authenticated.

Checkout Rules

Do not use git worktree in this workflow.

  • Stay in the user's current checkout.
  • Use local branches in the current repository when preparing the baseline branch and the PR review branch.
  • If uncommitted local changes make that unsafe, stop and tell the user instead of creating a worktree.

Workflow

  1. Identify the parent repo and its default branch. Use the actual remote default branch rather than assuming main or master.

  2. List all open pull requests on the parent repo. Prefer the parent repo over a fork unless the user says otherwise. Use gh pr list against the parent repo.

When considering candidates, skip PRs where you already requested changes or clarification in a previous next-pr cycle and the PR has not received a response since that comment, unless the user explicitly asks to revisit that PR anyway.

  1. Pick the most trivial PR and explain why it is the easiest yes/no decision. Prefer small, low-risk PRs such as:
  • dependency bumps with narrow blast radius
  • docs-only or config-only changes
  • tiny test fixes
  • one-commit PRs with limited file scope

Avoid PRs that:

  • change behavior across many files
  • mix refactors with fixes
  • need product or architecture decisions
  • are hard to evaluate because baseline validation is broken
  1. Establish a clean baseline before touching the PR branch. Update the local default branch from origin and validate that exact branch first, unless you are starting a new next-pr cycle immediately after finishing the previous one and the default branch tip is still the exact validated result of that previous cycle.

In that immediate follow-on case, reuse the test and benchmark run from the end of the previous cycle as the baseline for the next PR instead of rerunning the same baseline commands.

Run the repo's real validation commands, not substitutes. Include benchmarks when the repo has a benchmark workflow.

If baseline validation fails:

  • do not treat PR validation as authoritative
  • fix the baseline first if that matches the user's requested workflow
  • otherwise stop and tell the user the PR cannot be judged against a green baseline yet

If any of these changed since the previous cycle, do not reuse the old baseline:

  • the default branch tip
  • the validation command set
  • the local checkout or environment in a way that could affect results
  1. Check out the PR only after the baseline is green. Create a dedicated local review branch for the PR in the current checkout. Use gh pr checkout or an equivalent gh-driven fetch to materialize the PR locally. Rebase it on the latest default branch tip and collapse the PR changes into a single commit on top of that tip. Resolve conflicts carefully without discarding upstream or user work.

  2. Validate the collapsed rebased PR branch. Run the same tests and benchmarks used for the baseline so the comparison is meaningful. Call out:

  • passes or failures
  • benchmark regressions or improvements
  • new warnings, flaky behavior, or environment issues

If the PR fixes a Java bug and does not include a regression test:

  • add a regression test as part of the review branch before approval
  • verify that the regression test fails against the baseline branch without the fix
  • verify that the same regression test passes on the rebased PR branch with the fix
  1. Summarize for approval before merging. Report:
  • PR number and title
  • why it was chosen
  • what changed at a high level
  • rebase/collapse result
  • baseline validation result
  • PR validation result
  • recommendation and any risks

Do not merge or push yet.

If the review shows the PR is cosmetic or otherwise a no-op, and it does not fix a demonstrated bug or change behavior:

  • recommend closing it rather than merging it
  • explain briefly why it is a no-op
  • use gh pr close with a polite comment after the user approves closing it

If the review shows the PR is directionally useful but not ready to merge because it needs upstream fixes, more evidence, or clarification:

  • recommend requesting changes or clarification rather than merging or closing it
  • explain briefly what needs to change or what question needs to be answered
  • after the user approves that course, leave a short gh pr comment explaining the requested changes or clarification
  • treat that comment as the completion of the current cycle for that PR
  1. Merge only after explicit user approval. After approval, integrate the single collapsed PR commit into the default branch in the way the repo expects.

  2. Push only after the merge is complete and the default branch is in the intended state. Push the default branch to the appropriate remote and report what was pushed. If the pushed default branch tip matches the rebased PR branch you just validated, treat that validation run as the baseline for the next cycle.

  3. Close the PR manually if GitHub does not auto-close it after the merge push. When a PR was merged by rebasing or otherwise rewriting commits, verify whether it is still open. If it is, close it with gh pr close and leave a short comment explaining that the changes were merged onto the default branch via the collapsed rebased commit.

  4. Immediately start the workflow again on the next simplest remaining PR unless the user explicitly asked to stop after one PR. Go back to step 2, pick the next easiest remaining decision, and keep working through the queue. When step 9 left you on the same validated default branch tip, use that just-finished validation run as the baseline for the new cycle instead of rerunning baseline tests and benchmarks.

Show full SKILL.md (375 more words)Show less

If the previous cycle ended by requesting changes or clarification on a PR, skip that PR in later cycles until someone responds after your comment. Once there is a response, it becomes eligible again.

Validation Rules

  • Use the same command set for baseline and PR validation so results are comparable.
  • If the repo has both fast checks and full checks, run the full set unless the user explicitly narrows scope.
  • If benchmarks are noisy, report broad direction and uncertainty instead of overclaiming precision.
  • If a failure is clearly pre-existing, say that explicitly and separate it from PR-specific findings.
  • For Java bugfix PRs, require a regression test. If one is missing, add it on the review branch and prove it fails before the fix and passes after the fix.
  • If a PR is cosmetic or a no-op, do not merge it just because validation is green; require evidence of a real bugfix or behavior change.
  • If a PR needs upstream fixes or clarification, do not force it into a merge/close decision; requesting changes or clarification is a valid third outcome.
  • When chaining directly from one completed cycle to the next, you may reuse the just-validated default-branch state as the next baseline instead of rerunning baseline checks, but only if the branch tip and validation command set are unchanged and nothing else invalidated that result.

Communication

  • Tell the user which PR you picked and why it is the lowest-risk candidate.
  • When baseline is broken, say that first.
  • Before merge and push, stop at an approval gate and wait for the user.
  • After approval, report exactly which branch was updated and pushed.
  • If a rebased merge leaves the PR open, close it and say which comment or reason you used.
  • If closing a cosmetic or no-op PR, use a short polite message that explains it did not produce a demonstrated bugfix or behavior change and can be reopened with a concrete repro or regression test.
  • If requesting changes or clarification, say exactly what you asked for and that the PR will be skipped in later next-pr cycles until someone responds.
  • After completing one PR, say that you are immediately continuing to the next easiest remaining PR and whether you reused the previous cycle's validation as the new baseline.

© georgewfraser, MIT. 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 .agents/skills/next-pr of georgewfraser/java-language-server.

  • SKILL.md
  • agents/openai.yaml

Open the folder on GitHubat commit 58daaa2

Compare with similar skills

Next PR 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.

Next PR compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Next PR this skillgeorgewfraser/java-language-server814—~2.2kAutomated safety check: PassMIT
Finishing a Development Branchobra/superpowers297k5 repos~1.9kAutomated safety check: PassMIT
PR Babysitteropeninterpreter/openinterpreter69k3 repos~4.2kAutomated safety check: PassApache-2.0
Check PRonyx-dot-app/onyx32k2 repos~2.3kAutomated safety check: PassMIT
PR Design DocOpenHands/OpenHands90k—~2.4kAutomated safety check: PassMIT
WooCommerce Code Reviewwoocommerce/woocommerce11k3 repos~1.1kAutomated safety check: PassCustom licence

Similar skills

  • Walks the last step of a branch: confirm tests pass, detect the git environment, ask how to integrate, carry out your choice and clean up the worktree.

    297k GitHub starsUsed in 5 repos~1.9k tokens
    DevelopmentAuto-check passed
  • PR Babysitter

    openinterpreter/openinterpreter

    Watches an open GitHub pull request until it merges, handling review comments, diagnosing CI failures and retrying flaky checks along the way.

    69k GitHub starsUsed in 3 repos~4.2k tokens
    DevelopmentAuto-check passed
  • Check PR

    onyx-dot-app/onyx

    Checks a GitHub, GitLab, or Perforce (p4) pull request (or merge request, or shelved changelist) for unresolved review comments, failing status checks, and incomplete PR descriptions.

    32k GitHub starsUsed in 2 repos~2.3k tokens
    DevelopmentAuto-check passed
  • PR Design Doc

    OpenHands/OpenHands

    For a non-trivial pull request, write a self-contained HTML design doc under the temporary .pr/ directory and link a visibility-appropriate preview in the PR description, so maintainers grasp the…

    90k GitHub stars~2.4k tokensUpdated today
    DevelopmentAuto-check passed
  • WooCommerce Code Review

    woocommerce/woocommerce

    Reviews WooCommerce code changes against the project's standards, flagging backend PHP architecture, naming, documentation, data integrity and testing violations.

    11k GitHub starsUsed in 3 repos~1.1k tokens
    DevelopmentAuto-check passed
  • Record PR Demo

    payloadcms/payload

    A skill your agent uses when a Payload pull request needs a concise visual walkthrough for reviewers.

    45k GitHub stars~1k tokensUpdated yesterday
    DevelopmentAuto-check passed

Categories

Questions about Next PR

What does Next PR do?

Review and advance the next easy-to-merge pull request in this repository, then continue to the next simplest PR. Next PR is an agent skill from georgewfraser/java-language-server. Review and advance the next easy-to-merge pull request in this repository, then continue to the next simplest PR.

When should I use Next PR?

Next PR fits situations like: the user wants Codex to list open PRs; pick the most trivial low-risk PR; reuse a clean baseline on the default branch; collapse the rebased PR into a single commit on top of that branch.

How do I install Next PR in Claude Code?

Run `npx skills add georgewfraser/java-language-server --skill next-pr -a claude-code`. Or copy the skill folder (.agents/skills/next-pr in georgewfraser/java-language-server) into .claude/skills/next-pr in your project. Claude Code loads it when a task matches its description.

How do I install Next PR in Codex?

Run `npx skills add georgewfraser/java-language-server --skill next-pr -a codex`. Or copy the skill folder (.agents/skills/next-pr in georgewfraser/java-language-server) into .agents/skills/next-pr in your project. Codex loads it when a task matches its description.

Can I use Next PR 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 georgewfraser/java-language-server --skill next-pr -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/next-pr, .gemini/skills/next-pr, .github/skills/next-pr and .opencode/skills/next-pr in your project.

What does Next PR need to run?

Going by SKILL.md and its folder, Next PR needs the command-line tools its instructions call (gh and git).

Does Next PR access the network?

SKILL.md contains no URLs. Its commands use gh and git, which can reach the network depending on how they are called. This is read from the text; nothing was executed.

Is Next PR 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 Next PR use?

Next PR is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Next PR use?

About 2.2k tokens (SKILL.md is roughly 8.9k 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 Next PR?

Skills that share tags, products or a category with Next PR: Finishing a Development Branch (obra/superpowers, 297k stars), PR Babysitter (openinterpreter/openinterpreter, 69k stars), Check PR (onyx-dot-app/onyx, 32k stars) and PR Design Doc (OpenHands/OpenHands, 90k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Next PR?

georgewfraser (a GitHub user) maintains it in georgewfraser/java-language-server, which has 814 GitHub stars. The repository was last updated on October 2, 2026.

Source: georgewfraser/java-language-server on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.