# Create Pull Request > Steps for creating a pull request in this project. Use when this capability is needed. - Skill: `tomevault-io/create-pull-request-2` (Agent Skill, multi-file: 2 files) - Install (CLI): `npx skillmds@latest add tomevault-io/create-pull-request-2` - Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/create-pull-request-2/raw - Safety review: pending (external: skill-scanner PASS, skillspector PASS) - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Coding & Dev Tools - Author: tomevault-io (https://skillmd.com/u/tomevault-io) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/tomevault-io/create-pull-request-2 --- ## Requirements - GitHub CLI (`gh`) installed and authenticated - `gh pr create` — open the PR creation flow. - `gh pr view` — verify the PR was created correctly. - On a feature branch (not main) ## Steps - Infer from context or ask the user if this PR fixes or relates to an issue, follow up and ask for the issue number before proceeding. - Infer from context or ask the user to describe their AI involvement and review state: - AI generated the code and PR description — has it been reviewed by a human? - AI assisted with the PR description only — has it been reviewed by a human? - Other (ask user to describe) — has it been reviewed by a human? - Verify any needed docs and tests changes have been included. - Run appropriate quality assurance locally before opening the PR. ## PR title guidance - Use a descriptive title that communicates the problem solved or feature added in a few words. - Do not use vague titles like "Fixes #123" — write "Fix documentation dark mode refresh issue (#123)" instead. ## Template - If fully closes: Fixes #number | If related: Relates to #number Fixes/Relates to # ### Description - Describe the "why" of the changes. Concisely, why the proposed solution is the right one, and which other solutions might have been skipped. - Highlight areas of the proposed changes that require careful review. - Guidance on QA done so far vs. testing left to do. ### Testing - Summarise results briefly, e.g. "All tests pass, no lint errors" - Only report lint/test results for files included in this PR. Do not mention unrelated or untracked files. - Share guidance on any specific tests related to the PR - command(s) and results. - For frontend changes, add before/after screenshots if possible. - Note any testing still left to do or areas needing careful review ### AI usage - Agent: infer from your context / share user input. --- > Source: [wagtail/guide](https://github.com/wagtail/guide) — distributed by [TomeVault](https://tomevault.io).