SkillAgentSearch skills...

pr-review

Address feedback left on a GitHub pull request: fetch unresolved review threads, make agreed Elixir/Phoenix code fixes, reply, and resolve. Use for a PR URL/number or reviewer comments

Install / Use

npx skills add oliver-kriska/claude-elixir-phoenix --skill pr-review

Installs into whichever agent you are using.

About this skill
๐Ÿ“„

SKILL.md

Installable skill definition

Quality Score

88/100

Category

Automation

Supported Platforms

Universal

Our assessment of pr-review

pr-review scores 88/100 on our quality scale, 1471st of 2,768 Automation skills we index.

Its SKILL.md is 6.0 KB long, well organised into 11 sections with 3 code examples: a thorough specification that gives an agent plenty to work with.

It has 574 GitHub stars, a meaningful sign that others use it.

Substance
29/30
Structure
18/20
Description
15/15
Adoption
12/20
Freshness
15/15

Maintenance, license and trust

  • The repository was last updated 12 days ago, so pr-review is actively maintained.
  • Our last check on 2026-10-31 found the source still online.
  • It is released under the MIT license, a permissive license that allows use, modification and commercial use with attribution.
  • Its trust signals score 100/100, with no cautions. These come from repository metadata, not a code audit โ€” read the skill file before letting an agent act on it.

Safety scan

No issues found

Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful.

AI review by kimi-k2.7-code on 2026-10-04. Automated pattern scan on 2026-10-04. It catches known dangerous patterns, not every risk โ€” read a skill before letting an agent act on it.

pr-review compared with similar skills

All 4 of these similar skills score higher than pr-review; compare them before choosing.

SkillScoreStarsUpdatedFormat
pr-review (this skill)by oliver-kriska8857412d agoSKILL.md
Agent-Reachby Panniantong100113.0k10d agoCLAUDE.md
rufloby ruvnet10082.0ktodayCLAUDE.md
Claude-Code-Game-Studiosby Donchitos10040.1k9d agoCLAUDE.md
Scraplingby D4Vinci100104.2k1d agoMCP Server

Frequently asked questions

How do I install pr-review?
Run npx skills add oliver-kriska/claude-elixir-phoenix --skill pr-review. The install tabs above show the steps for each supported agent.
Which AI agents does pr-review work with?
It is written for Universal, as a SKILL.md file. Other agents that read the same format can often use it too.
Is pr-review safe to use?
Our scan of the whole file found no instruction hijacking, hidden characters, credential access, data exfiltration or destructive commands. An AI review of the same text found nothing harmful. It is MIT-licensed and scores 100/100 on trust signals. Skills are instructions an agent will follow, so read the file before installing it and do not approve commands you do not understand.
Is pr-review still maintained?
The repository was last updated 12 days ago, so pr-review is actively maintained.

name: pr-review description: "Address feedback left on a GitHub pull request: fetch unresolved review threads, make agreed Elixir/Phoenix code fixes, reply, and resolve. Use for a PR URL/number or reviewer comments. NOT for pre-PR review, findings triage, or CI monitoring." effort: high argument-hint: <PR number or URL> [--fix] [--bots-only] [--no-resolve]

PR Review Response

Close the review loop: fetch unresolved threads โ†’ fix โ†’ reply โ†’ resolve. GitHub's isResolved is the state โ€” re-runs are idempotent, handled threads drop out automatically.

Usage

/phx:pr-review 42                  # Triage unresolved threads on PR #42
/phx:pr-review 42 --fix            # Triage + apply approved code fixes
/phx:pr-review https://...         # Full URL also works (repo parsed from URL)
/phx:pr-review 42 --bots-only      # Triage only CI bot threads (Copilot, Codex...)
/phx:pr-review 42 --no-resolve     # Reply but leave threads open

Step 1: Resolve PR + Fetch Threads

gh pr view "$PR" --json number,title,state,baseRefName,headRefName,url,author (accepts number or URL; URL also yields owner/repo). Then fetch ALL review threads with thread IDs + resolved status โ€” REST alone cannot do this:

cat > /tmp/review_threads.graphql <<'GQL'
query($owner:String!, $repo:String!, $pr:Int!, $cursor:String) {
  repository(owner:$owner, name:$repo) {
    pullRequest(number:$pr) {
      reviewThreads(first:50, after:$cursor) {
        pageInfo { hasNextPage endCursor }
        nodes {
          id isResolved isOutdated path line originalLine
          comments(first:20) { nodes {
            databaseId body createdAt
            author { login __typename } } }
        }
      }
    }
  }
}
GQL
gh api graphql --paginate -F owner="$OWNER" -F repo="$REPO" -F pr="$PR" \
  -F query=@/tmp/review_threads.graphql \
  --jq '.data.repository.pullRequest.reviewThreads.nodes[]
        | select(.isResolved == false)
        | {threadId: .id, isOutdated, path, line: (.line // .originalLine),
           firstCommentId: .comments.nodes[0].databaseId,
           author: .comments.nodes[0].author.login,
           isBot: (.comments.nodes[0].author.__typename == "Bot"),
           body: .comments.nodes[0].body}'

Also fetch review summaries (gh api "repos/$OWNER/$REPO/pulls/$PR/reviews") โ€” they are NOT threads and cannot be resolved; surface CHANGES_REQUESTED bodies separately. Bot detection: __typename == "Bot" / user.type == "Bot" (the [bot] login suffix is NOT reliable across endpoints).

Step 2: Triage Table

Group by file, one row per thread. With --bots-only, keep only isBot rows.

| # | file:line | author | category | proposed action | |---|-----------|--------|----------|-----------------|

Categories: code-change ("should be", "use X instead") ยท question ("why", "how does") ยท nitpick ("nit:", style) ยท praise (no action) ยท discussion (architecture) ยท bot-finding (CI bot inline comment โ€” verify before accepting, many are false positives) ยท outdated (isOutdated: true โ€” line moved; default: reply "addressed in {commit}" + resolve). Present the table and let the user greenlight threads.

Step 3: Per-Thread Loop

For each greenlit thread:

  1. Read code at path:line; check the suggestion against Iron Laws

  2. Apply fix with a user-visible diff (only with --fix or explicit ok)

  3. Draft reply (templates: ${CLAUDE_SKILL_DIR}/references/response-patterns.md)

  4. STOP โ€” show diff + reply, get confirmation

  5. Post reply โ€” REST, targeting the thread's root comment:

    gh api --method POST \
      "repos/$OWNER/$REPO/pulls/$PR/comments/$FIRST_COMMENT_ID/replies" \
      -f body="$REPLY_TEXT"
    
  6. Resolve the thread (skip with --no-resolve):

    gh api graphql -f query='mutation($threadId:ID!){
      resolveReviewThread(input:{threadId:$threadId}){
        thread { id isResolved } }}' -F threadId="$THREAD_ID"
    

Mistake recovery: unresolveReviewThread takes the same input shape.

Step 4: Verify

mix compile --warnings-as-errors && mix test scoped to changed files. Do NOT commit or push โ€” leave that to the user.

Step 5: Final Summary

Print rollup: # | thread | action | status (replied/resolved/skipped). List changed files. Optionally post a top-level conversation comment (gh api --method POST "repos/$OWNER/$REPO/issues/$PR/comments" -f body=...) with the rollup โ€” only on user approval.

Iron Laws

  1. NEVER auto-post responses โ€” Always show drafts and get explicit approval
  2. NEVER dismiss a review โ€” Only the reviewer should dismiss
  3. Iron Laws override reviewer suggestions โ€” If a suggestion violates an Iron Law, explain why in the reply
  4. Keep responses constructive โ€” Acknowledge the feedback, explain reasoning
  5. Separate fixes from responses โ€” Apply code changes in a distinct step
  6. NEVER resolve a thread without first posting a reply โ€” every resolve is preceded by a reply on that thread explaining what was done
  7. NEVER claim a fix without a shown diff โ€” no "should be fixed" replies without a user-visible change
  8. Bot findings get the same scrutiny as humans โ€” decline Iron-Law-violating bot suggestions with explanation; never bulk-resolve "bot noise" without replies

Integration

PR receives review โ†’ /phx:pr-review {number}  โ† YOU ARE HERE
   โ†“ fetch unresolved threads (GraphQL, paginated)
   โ†“ triage table โ†’ user greenlights
   โ†“ per thread: fix (diff) โ†’ reply โ†’ resolve
   โ†“ verify (mix compile + test) โ†’ summary
Push changes โ†’ user handles git push

Next Steps

  • /phx:plan โ€” if findings reveal scope gaps
  • /phx:verify โ€” full verification before pushing
  • Re-run /phx:pr-review after the next review round (idempotent)

References

  • ${CLAUDE_SKILL_DIR}/references/response-patterns.md โ€” Response templates and tone
  • ${CLAUDE_SKILL_DIR}/references/gh-commands.md โ€” Full gh command reference (3 comment surfaces, pagination, bot detection)
  • ${CLAUDE_SKILL_DIR}/references/bot-triage.md โ€” Batch-triaging CI bot review passes

Related Skills

View on GitHub
GitHub Stars574
CategoryAutomation
Updated12d ago
Forks45

Languages

Python

Trust signals

100/100

From repository metadata: license, adoption, age and documentation. Not a code audit โ€” see the Safety scan above for what the skill file itself contains.

No cautions
pr-review โ€” Universal Skill: Install & Safety Check | SkillAgent