worktree_pr — how to use worktree_pr how to use worktree_pr, worktree_pr setup guide, FAIR data on HDF5, autonomous pull request creation, git branch automation, worktree_pr vs pr_create, worktree_pr install, what is worktree_pr, worktree_pr alternative

v1.0.0
GitHub

About this Skill

Perfect for DevOps Agents needing autonomous pull request creation for scientific data products in HDF5 format. worktree_pr is a skill that automates the creation of pull requests using FAIR data on HDF5, following a self-describing data format for scientific data products.

Features

Automates pull request creation using git branch and issue data
Generates PR text from commits and issues without user feedback
Follows branch naming convention using <type>/<issue_number>-<summary> format
Runs using git branch --show-current command
Creates PRs based on issue branch data, matching specific naming conventions

# Core Topics

vig-os vig-os
[0]
[2]
Updated: 2/24/2026

Quality Score

Top 5%
39
Excellent
Based on code quality & docs
Installation
SYS Universal Install (Auto-Detect)
Cursor IDE Windsurf IDE VS Code IDE
> npx killer-skills add vig-os/fd5/worktree_pr

Agent Capability Analysis

The worktree_pr MCP Server by vig-os is an open-source Categories.community integration for Claude and other AI agents, enabling seamless task automation and capability expansion. Optimized for how to use worktree_pr, worktree_pr setup guide, FAIR data on HDF5.

Ideal Agent Persona

Perfect for DevOps Agents needing autonomous pull request creation for scientific data products in HDF5 format.

Core Value

Empowers agents to automate pull requests without user interaction, utilizing FAIR-principled data formats and generating PR text from commits and issues, leveraging Git branch naming conventions and HDF5 data format.

Capabilities Granted for worktree_pr MCP Server

Automating pull requests for scientific data updates
Generating PR text from commit messages and issue trackers
Streamlining data product deployment using FAIR principles

! Prerequisites & Limits

  • Requires Git repository access
  • Adherence to specific branch naming conventions
  • HDF5 data format compatibility
Project
SKILL.md
5.3 KB
.cursorrules
1.2 KB
package.json
240 B
Ready
UTF-8

# Tags

[No tags]
SKILL.md
Readonly

Autonomous PR

Create a pull request without user interaction. This is the worktree variant of pr_create.

Rule: no blocking for feedback. Auto-generate PR text from commits and issue.

Precondition: Issue Branch Required

  1. Run: git branch --show-current
  2. The branch name must match <type>/<issue_number>-<summary> (e.g. feature/79-declarative-sync-manifest). See branch-naming.mdc for the full convention.
  3. Extract the <issue_number> from the branch name.

Workflow Steps

1. Ensure clean state

bash
1git status 2git fetch origin
  • If there are uncommitted changes, commit them first.
  • Push the branch: git push -u origin HEAD

2. Determine base branch

Detect whether this issue is a sub-issue and resolve the correct merge target:

  1. Determine the repo: gh repo view --json nameWithOwner --jq '.nameWithOwner'

  2. Check for a parent issue:

    bash
    1gh api repos/{owner}/{repo}/issues/{issue_number}/parent --jq '.number'
  3. If a parent exists, resolve its linked branch:

    bash
    1gh issue develop --list <parent_number>
    • Use the parent's branch as <base_branch>.
    • If the parent has no linked branch, fall back to dev.
  4. If no parent exists, use dev as <base_branch>.

3. Gather context

bash
1git log <base_branch>..HEAD --oneline 2git diff <base_branch>...HEAD --stat 3gh issue view <issue_number> --json title,body
  • Read the issue title and acceptance criteria.
  • Summarize what the commits accomplish.

4. Ensure CHANGELOG is updated

  • Check CHANGELOG.md for an entry under ## Unreleased that covers the changes.
  • If missing, add the appropriate entry and commit.

5. Generate PR text

  1. Read the template: cat .github/pull_request_template.md
  2. Use it as the literal skeleton — keep every heading, every checkbox line, every sub-heading. Strip only the HTML comments (<!-- ... -->).
  3. Section-by-section mapping:
    • Description: Summarize what the PR does from the issue body and commit messages.
    • Type of Change: Check the single box matching the branch type / commit types. Check Breaking change modifier only if commits contain !.
    • Changes Made: List changed files with bullet sub-details (from git diff --stat and git log).
    • Changelog Entry: Paste the exact ## Unreleased diff from CHANGELOG.md. If no changelog update, write "No changelog needed" and explain.
    • Testing: Check Tests pass locally if tests were run. Check Manual testing performed only if actually done. Fill Manual Testing Details or write "N/A".
    • Checklist: Check only items that are genuinely true. Leave unchecked items unchecked — do not remove them.
    • Additional Notes: Add design links, context, or write "N/A".
    • Refs: Refs: #<issue_number>
  4. Explicit prohibitions: Do not invent new sections. Do not rename headings. Do not omit sections. Do not remove unchecked boxes.
  5. Write the body to .github/pr-draft-<issue_number>.md.

6. Create PR

bash
1# Append reviewer if PR_REVIEWER is set in environment 2REVIEWER_ARG="" 3if [ -n "${PR_REVIEWER:-}" ]; then 4 REVIEWER_ARG="--reviewer $PR_REVIEWER" 5fi 6 7gh pr create --base <base_branch> --title "<type>: <description> (#<issue_number>)" \ 8 --body-file .github/pr-draft-<issue_number>.md \ 9 --assignee @me $REVIEWER_ARG

If the WORKTREE_REVIEWER environment variable is set (populated by just worktree-start), add the reviewer:

bash
1gh pr create --base <base_branch> --title "<type>: <description> (#<issue_number>)" \ 2 --body-file .github/pr-draft-<issue_number>.md \ 3 --assignee @me \ 4 --reviewer "$WORKTREE_REVIEWER"

The reviewer is the person who launched the worktree (their gh user login), not the agent.

7. Clean up

  • Delete the draft file: rm .github/pr-draft-<issue_number>.md
  • Report the PR URL.

Delegation

The following steps SHOULD be delegated to reduce token consumption:

  • Steps 1-2 (precondition check, ensure clean state, determine base branch): Spawn a Task subagent with model: "fast" that validates the branch name, runs git status/git fetch, pushes the branch, checks for a parent issue via gh api, resolves the base branch. Returns: issue number, base branch name, clean state confirmation.
  • Step 3 (gather context): Spawn a Task subagent with model: "fast" that executes git log, git diff, gh issue view and returns the raw outputs. Returns: commit log, diff stat, issue title/body.
  • Steps 6-7 (create PR, clean up): Spawn a Task subagent with model: "fast" that takes the PR title and body file path, executes gh pr create, deletes the draft file, and returns the PR URL.

Steps 4-5 (ensure CHANGELOG updated, generate PR text) should remain in the main agent as they require understanding changes and writing structured content.

Reference: subagent-delegation rule

Important Notes

  • Never block for user review of the PR text. Generate the best text from available context.
  • Base branch is auto-detected: parent issue's branch for sub-issues, dev otherwise.
  • The PR title should follow commit message conventions: type(scope): description (#issue).
  • NEVER add 'Co-authored-by: Cursor cursoragent@cursor.com' to commit messages.

Related Skills

Looking for an alternative to worktree_pr or building a Categories.community AI Agent? Explore these related open-source MCP Servers.

View All

widget-generator

Logo of f
f

widget-generator is an open-source AI agent skill for creating widget plugins that are injected into prompt feeds on prompts.chat. It supports two rendering modes: standard prompt widgets using default PromptCard styling and custom render widgets built as full React components.

149.6k
0
Design

chat-sdk

Logo of lobehub
lobehub

chat-sdk is a unified TypeScript SDK for building chat bots across multiple platforms, providing a single interface for deploying bot logic.

73.0k
0
Communication

zustand

Logo of lobehub
lobehub

The ultimate space for work and life — to find, build, and collaborate with agent teammates that grow with you. We are taking agent harness to the next level — enabling multi-agent collaboration, effortless agent team design, and introducing agents as the unit of work interaction.

72.8k
0
Communication

data-fetching

Logo of lobehub
lobehub

The ultimate space for work and life — to find, build, and collaborate with agent teammates that grow with you. We are taking agent harness to the next level — enabling multi-agent collaboration, effortless agent team design, and introducing agents as the unit of work interaction.

72.8k
0
Communication