Ready to supercharge your AI agents?

Install any skill in seconds. Free to start, no credit card required.

Get Started Free →

Pull Request Manager Agent · ruvnet | DecimalAI

Agent Github PR Manager

SkillScore13Maintained

agent-github-pr-manager

Agent skill for github-pr-manager - invoke with $agent-github-pr-manager

by ruvnet

Links this skill to your agents and keeps it current with the author’s updates.

Automatic

Coding pack

1 · One-time — install the CLI (docs ↗)

pip install decimalai

2 · Get the skill — writes .claude/skills/agent-github-pr-manager/SKILL.md

decimalai skills pull agent-github-pr-manager --out .claude/skills/

No CLI? Get the same file — curl · raw file

curl -fsSL --create-dirs -o .claude/skills/agent-github-pr-manager/SKILL.md https://app.decimal.ai/s/agent-github-pr-manager/SKILL.md

Raw file: /s/agent-github-pr-manager/SKILL.md ↗ · machine-readable summary at /s/agent-github-pr-manager.json ↗

In productionGlobal

how this skill behaves on real traffic · last 30 days

No production data for this skill yet — usage numbers appear here once agents start running it.

Anonymized aggregate across every org using this skill — proof from real production use, not a demo.

Does this skill help? · Benchmark: with vs. withoutsign in to run this on your agents

23 test cases from the publisher’s eval.yaml · each run twice on gemini-3.6-flash — with the skill injected, then without · last run 61d ago

Yes — +9% lift · at higher cost+31% tokens0% turns13% of whole cases pass with skill (3/23)

Test case	Without → With	Effect	Δ tokens	Δ turns
case-09	✗→✓	▲ Improved	—	—
case-08	✗→✓	▲ Improved	—	—
case-13	✓→✓	＝ Same ✓	—	—
case-20	✗→✗	＝ Same ✗	—	—
case-02	✗→✗	＝ Same ✗	—	—

Showing 5 of 23 cases

SkillSafety

✓ 3 of 3 passed

Code scan15 of 15 checks clean

0 critical · 0 warning · 0 info — scanner v4 — 7/30/2026

live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing

IntentSafe

LLM judge · does the skill do what it claims

ContentClean

LLM judge · policy & harmful-content review

Source

Imported

ruvnet / ruflo

.agents/skills/agent-github-pr-manager/SKILL.md

View on GitHub ↗

License MITImported 8/2/2026Version v1

Skill contents

SKILL.md

Pull Request Manager Agent

Purpose

This agent specializes in managing the complete lifecycle of pull requests, from creation through review to merge, using GitHub's gh CLI and swarm coordination for complex workflows.

Core Functionality

1. PR Creation & Management

Creates PRs with comprehensive descriptions
Sets up review assignments
Configures auto-merge when appropriate
Links related issues automatically

2. Review Coordination

Spawns specialized review agents
Coordinates security, performance, and code quality reviews
Aggregates feedback from multiple reviewers
Manages review iterations

3. Merge Strategies

Squash: For feature branches with many commits
Merge: For preserving complete history
Rebase: For linear history
Handles merge conflicts intelligently

4. CI/CD Integration

Monitors test status
Ensures all checks pass
Coordinates with deployment pipelines
Handles rollback if needed

Usage Examples

Simple PR Creation

"Create a PR for the feature$auth-system branch"

Complex Review Workflow

"Create a PR with multi-stage review including security audit and performance testing"

Automated Merge

"Set up auto-merge for the bugfix PR after all tests pass"

Workflow Patterns

1. Standard Feature PR

bash
1. Create PR with detailed description
2. Assign reviewers based on CODEOWNERS
3. Run automated checks
4. Coordinate human reviews
5. Address feedback
6. Merge when approved

2. Hotfix PR

bash
1. Create urgent PR
2. Fast-track review process
3. Run critical tests only
4. Merge with admin override if needed
5. Backport to release branches

3. Large Feature PR

bash
1. Create draft PR early
2. Spawn specialized review agents
3. Coordinate phased reviews
4. Run comprehensive test suites
5. Staged merge with feature flags

GitHub CLI Integration

Common Commands

bash
# Create PR
gh pr create --title "..." --body "..." --base main

# Review PR
gh pr review --approve --body "LGTM"

# Check status
gh pr status --json state,statusCheckRollup

# Merge PR
gh pr merge --squash --delete-branch

Multi-Agent Coordination

Review Swarm Setup

Initialize review swarm
Spawn specialized agents:

Code quality reviewer
Security auditor
Performance analyzer
Documentation checker

Coordinate parallel reviews
Synthesize feedback

Integration with Other Agents

Code Review Coordinator: For detailed code analysis
Release Manager: For version coordination
Issue Tracker: For linked issue updates
CI/CD Orchestrator: For pipeline management

Best Practices

PR Description Template

markdown
## Summary
Brief description of changes

## Motivation
Why these changes are needed

## Changes
- List of specific changes
- Breaking changes highlighted

## Testing
- How changes were tested
- Test coverage metrics

## Checklist
- [ ] Tests pass
- [ ] Documentation updated
- [ ] No breaking changes (or documented)

Review Coordination

Assign domain experts for specialized reviews
Use draft PRs for early feedback
Batch similar PRs for efficiency
Maintain clear review SLAs

Error Handling

Common Issues

Merge Conflicts: Automated resolution for simple cases
Failed Tests: Retry flaky tests, investigate persistent failures
Review Delays: Escalation and reminder system
Branch Protection: Handle required reviews and status checks

Recovery Strategies

Automatic rebase for outdated branches
Conflict resolution assistance
Alternative merge strategies
Rollback procedures

Illustrative Want to see what agent-github-pr-manager would put in front of your model? Preview the rendered body and check its trigger phrases — no LLM call, no signup.

Latest benchmark · publisher runVerified

7/28/2026, 14:14 UTC · 23 cases

Lift

+9%

13% vs 4%

Agent turns

0%

1.0 vs 1.0

Tokens

+31%

1,885.78 vs 1,444.65

fail → pass · 2pass → pass · 1fail → fail · 20

Per-case results (23)

Case	Status	Duration (ms)			Turns			Tokens			Tool calls
Case	Status	Without	With	Δ	Without	With	Δ	Without	With	Δ	Without	With	Δ
case-20 If a newly merged pull request causes immediate failures in deployment verification, what safety procedure must be executed?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-02 I am querying pull request status in GitHub CLI (`gh`) and need JSON formatted output containing both overall PR state and status check details. Which specific `--json` fields must be requested?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-16 When merging a pull request into a protected branch with mandatory check requirements, how should branch protection rules be managed?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-01 I want to merge a pull request using the GitHub CLI tool. I need to consolidate all feature commits into a single commit and automatically remove the remote feature branch upon completion. What command and flags accomplish this?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-21 I am configuring GitHub branch protection settings using the REST API. Which REST endpoint URL and HTTP method are used to set required status checks and review requirements on a branch?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-13 A pull request branch has fallen behind target `main` and encountered simple merge conflicts. What automated recovery step should be performed?	pass→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-12 When starting a complex, multi-week feature implementation, what PR state should be used initially and what merge mechanism ensures safe rollout?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-17 When opening a pull request meant to address issue #89, how should issue tracking integration be handled during PR creation?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-22 I need to write a `.github/workflows/ci.yml` GitHub Actions pipeline that triggers on `pull_request` events to run `npm test`. Please provide the raw YAML file content.	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-04 When creating a new pull request with GitHub CLI (`gh`), I want to explicitly set `main` as the target destination branch. Which CLI parameter flag defines the destination branch?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-14 During CI check monitoring, how should an automated PR manager distinguish handling between intermittent test failures and persistent failures?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-19 When multiple agents in a swarm submit separate security, performance, and style notes on a single PR, how should these outputs be structured before presenting them?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-15 When a requested reviewer does not provide feedback within the required time window, what automated mechanism prevents PR bottlenecks?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-11 When handling an urgent production hotfix PR, what adjustments are made to test execution depth and post-merge branch maintenance?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-09 We are setting up a specialized multi-agent review swarm for auditing incoming pull requests. What four specialized domain review roles comprise this swarm?	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-03 I need to issue an official approval on a GitHub pull request via the GitHub CLI tool (`gh`) while adding the comment 'LGTM' in a single CLI command. What is the syntax?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-07 Our project enforces a completely flat linear git history on the main branch without any merge commit nodes. Which merge strategy re-applies branch commits directly onto main?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-23 I ran `git merge main` locally in my terminal and hit a conflict in `src/app.js` with standard conflict markers (`<<<<<<< HEAD`, `=======`, `>>>>>>> main`). How do I edit and stage this file to complete the local git merge?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-05 A contributor submitted a feature branch with 40 granular work-in-progress commits. We want to combine these into main as a single clean commit. Which merge strategy is appropriate?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-06 Our enterprise repository policy strictly requires retaining every individual commit and parent reference when integrating release branches into main. Which merge strategy fits this requirement?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-08 I am writing a standardized GitHub pull request template file. What are the five required primary markdown headings included in the standard PR template structure?	fail→pass	—	—	—	—	—	—	—	—	—	—	—	—
case-10 In a standard feature PR workflow, how should the initial reviewer assignment be automatically decided upon PR creation?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—
case-18 We want a pull request to automatically complete as soon as CI passes without manual intervention. At what point in the PR lifecycle is auto-merge configured?	fail→fail	—	—	—	—	—	—	—	—	—	—	—	—

These cases were also screened by the skill’s intent / safety pass.See the Trust review ↗Reproduce this benchmark in Colab ↗

How this skill was measured

DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted. The headline lift of +9 percentage points is the difference between those two pass rates over the 23 comparable cases.

Model measured: gemini-3.6-flash
Graded by: gemini-3.6-flash
Grading method: judged
Verification: verified
Pass rate without the skill: 4.4%
Pass rate with the skill: 13%
Cases attempted: 23
Cases counted toward lift: 23
Cases that regressed: 0
Skill version measured: v1
Run completed: 7/28/2026, 14:14 UTC

The per-case answers from this run were removed by the retention sweep, so the case table below shows the verdicts without the text either arm produced. The counts above were recorded at the time and are unaffected. Answers are now kept for 180 days.

SkillScore blends 4 signals1 of 4 complete

BenchmarkPass rate WITH skill on a verified run13

Live evalNeeds ≥10 evaluated traces in 30 days

AI ratingNeeds ≥10 rated traces

AdoptionNeeds ≥10 serves across ≥2 orgs in 30 days

Efficiency · separate axis (at equal correctness)+0% turns+4% tokens

Shown for transparency; not credited unless a verified run shows no correctness regression.

Backed by 1 of 4 signals — each added signal makes the score more robust.

Verified on gemini-3.6-flash · re-tested 2 months ago

Other measured skills in the registry, with their headline benchmark lift.