agentleFS
rlajousrlajous/claude-code-commands/skills/start/SKILL.md

start

Skill30 starsChanged 3 months agoMITView on GitHub

Start a new PR by creating a feature branch following repo conventions

130 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use setup-engine, team-audio and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • CoralGameSkill

    setup-engine

    Configure the project's game engine and version. Pins the engine in CLAUDE.md, detects knowledge gaps, and populates engine reference docs via WebSearch when the version is beyond the LLM's training data.

    Also in 13 projects that use this

  • CoralGameSkill

    team-audio

    Orchestrate audio team: audio-director + sound-designer + technical-artist + gameplay-programmer for full audio pipeline from direction to implementation.

    Also in 13 projects that use this

  • lilangMaxSkill

    create-stories

    Break a single epic into implementable story files. Reads the epic, its GDD, governing ADRs, and control manifest. Each story embeds its GDD requirement TR-ID, ADR guidance, acceptance criteria, story type, and test evidence path. Run after /create-epics for each epic.

    Also in 11 projects that use this

  • lilangMaxSkill

    story-done

    End-of-story completion review. Reads the story file, verifies each acceptance criterion against the implementation, checks for GDD/ADR deviations, prompts code review, updates story status to Complete, and surfaces the next ready story from the sprint.

    Also in 11 projects that use this

  • +5 more from CoralGame
  • +3 more from lilangMax

The file

Show its rules and text

We split this skill into the rules it gives an agent

15 rules. 0 are common, 9 are distinctive, and 6 are very specific to this project.

  • 0

    Common

    Most of the 117 edited versions we read keep these.

  • 9

    Distinctive

    This author's own choices. Worth a look.

  • 6

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Ask the user for the following information (use the active host user-input mechanism):

    line 54
  2. Very specific:

    If commits.requireTicket: true in config, ticket is required

    line 76
  3. Distinctive:

    If no ticket provided and not required, proceed without it

    line 78
  4. Distinctive:

    If a ticket ID was provided and issue tracker is configured:

    line 104
  5. Distinctive:

    If Linear ticket format is detected, use the Linear MCP server:

    line 116
  6. Distinctive:

    If Jira is configured, use the Jira MCP server:

    line 127
  7. Distinctive:

    If GitHub format detected:

    line 138
  8. Distinctive:

    If branch already exists, ask: "Branch exists.

    line 199
  9. Distinctive:

    If not on expected base branch, warn and confirm

    line 200
  10. Very specific:

    Create .git-workflow/pr-context.json with collected information:

    line 204
  11. Very specific:

    Create .git-workflow/ before writing.

    line 229
  12. Very specific:

    If .git-workflow/pr-context.json exists, ask before overwriting.

    line 230
  13. Very specific:

    Never overwrite or delete the legacy .claude/.pr-context.json fallback.

    line 231
  14. Distinctive:

    If ticket details were fetched, include:

    line 248
  15. Very specific:

    All configurable options from .git-workflow/config.yaml:

    line 258

“Common” is measured: kept by at least 6 in 10 of the 117 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Cross-runtime: follow runtime compatibility for invocation, delegation, configuration precedence, state paths, and permissions.

You are helping create a new pull request. Your task is to create a properly named feature branch, optionally fetch ticket details, and store context for subsequent commands.

Step 1: Load Configuration

Resolve project configuration with canonical-first precedence:

if [ -f ".git-workflow/config.yaml" ]; then
  CONFIG_PATH=".git-workflow/config.yaml"
elif [ -f ".claude/config.yaml" ]; then
  CONFIG_PATH=".claude/config.yaml" # legacy read-only fallback
else
  CONFIG_PATH=""
fi

Configuration Priority:

  1. .git-workflow/config.yaml (if exists)
  2. .claude/config.yaml (legacy read-only fallback)
  3. Auto-detection (package.json, pyproject.toml, etc.)
  4. Sensible defaults

Default Values (when no config):

workflow:
  type: staging
  developmentBranch: staging
  productionBranch: main
branches:
  feature: "{type}/{ticket}-{description}"
commits:
  types: [Feature, Fix, Hotfix, Refactor, Docs, Test, Chore]
  requireTicket: false
  ticketPattern: "^[A-Z]+-\\d+$"
issueTracker:
  type: auto

Step 2: Gather Information

Ask the user for the following information (use the active host user-input mechanism):

Question 1: Ticket ID

Question: "What is the ticket ID or URL?"

  • Examples:
    • PROJ-1234 (Jira format)
    • ENG-456 (Linear format)
    • #789 (GitHub issue)
    • https://linear.app/team/issue/ENG-456/...
    • https://your-org.atlassian.net/browse/PROJ-1234

Parsing Logic:

  • Extract ticket ID from URL if provided
  • Linear: Pattern /issue/([A-Z]+-\d+)/
  • Jira: Pattern /browse/([A-Z]+-\d+)
  • GitHub: Pattern issues/(\d+) or #(\d+)

Validation:

  • If commits.requireTicket: true in config, ticket is required
  • Validate against commits.ticketPattern if provided
  • If no ticket provided and not required, proceed without it

Question 2: Change Type

Question: "What type of change is this?"

Options (from config or defaults):

  • feature - New feature or enhancement
  • fix - Bug fix
  • hotfix - Urgent production fix
  • chore - Maintenance, dependencies, etc.
  • refactor - Code refactoring
  • docs - Documentation changes

Question 3: Short Description

Question: "Provide a brief description (will be converted to kebab-case)"

Examples:

  • "add user authentication" -> add-user-authentication
  • "fix memory leak in parser" -> fix-memory-leak-in-parser

Step 3: Fetch Ticket Details (Optional)

If a ticket ID was provided and issue tracker is configured:

Auto-Detect Issue Tracker Type

Based on ticket format and available MCP servers:

  • ^[A-Z]+-\d+$ with Linear MCP available -> Linear
  • ^[A-Z]+-\d+$ with Jira config/MCP available -> Jira
  • ^#?\d+$ or GitHub URL -> GitHub Issues (via gh CLI)

Linear Integration

If Linear ticket format is detected, use the Linear MCP server:

# Fetch ticket details via Linear MCP
mcp__linear__get_issue(id: ticketId)

The MCP server handles authentication automatically via OAuth.

Jira Integration

If Jira is configured, use the Jira MCP server:

# Fetch ticket details via Jira MCP
mcp__jira__get_issue(issueKey: ticketId)

The MCP server handles authentication automatically. The jira.baseUrl from config is used to identify the Jira instance.

GitHub Issues

If GitHub format detected:

# Use gh CLI to fetch issue
gh issue view {issue_number} --json title,body,state

Store Fetched Data:

{
  "ticket_title": "Fetched title from issue tracker",
  "ticket_description": "Description from issue tracker",
  "ticket_status": "In Progress"
}

Step 4: Create Branch

Generate Branch Name

Apply the branch pattern from config (default: {type}/{ticket}-{description}):

Pattern: {type}/{ticket}-{description}
Type: fix
Ticket: proj-1234
Description: add-user-auth

Result: fix/proj-1234-add-user-auth

Branch Naming Rules:

  • Convert ticket to lowercase
  • Convert description to kebab-case (lowercase, hyphens)
  • Remove special characters
  • Limit total length to 100 characters

Examples:

  • fix/proj-1234-standardize-token-info-schema
  • feature/eng-456-add-holder-analysis
  • hotfix/abc-789-fix-critical-bug
  • chore/update-dependencies (no ticket)

Create and Checkout Branch

Immediately before running any command below, show the resolved base branch, pull target, new branch name, and canonical context destination. Ask for explicit confirmation to create and check out the branch. Do not treat automatic skill invocation or earlier information-gathering answers as approval.

# Ensure we're on the development branch and up-to-date
git checkout {workflow.developmentBranch} 2>/dev/null || git checkout main
git pull

# Create feature branch
git checkout -b {branch_name}

Error Handling:

  • If branch already exists, ask: "Branch exists. Switch to it or create a new name?"
  • If not on expected base branch, warn and confirm

Step 5: Store Context

Create .git-workflow/pr-context.json with collected information:

mkdir -p .git-workflow
{
  "ticket_id": "PROJ-1234",
  "ticket_url": "https://issue-tracker.com/PROJ-1234",
  "ticket_title": "Title from issue tracker (if fetched)",
  "branch": "fix/proj-1234-description",
  "type": "fix",
  "description": "Human readable description",
  "started_at": "2025-01-17T12:00:00Z",
  "config": {
    "workflow": "staging",
    "developmentBranch": "staging",
    "productionBranch": "main"
  }
}

Important:

  • Create .git-workflow/ before writing.
  • If .git-workflow/pr-context.json exists, ask before overwriting.
  • Never overwrite or delete the legacy .claude/.pr-context.json fallback.

Step 6: Confirm

Output a confirmation message:

Created branch: {branch_name}
Ticket: {ticket_id} {ticket_title if fetched}
Type: {type}

Next steps:
1. Make your code changes
2. Run /commit to stage and commit your changes
3. Run /finish to create the pull request

If ticket details were fetched, include:

Ticket Details:
- Title: {ticket_title}
- Status: {ticket_status}

Configuration Reference

All configurable options from .git-workflow/config.yaml:

SettingDefaultDescription
workflow.developmentBranchstagingBase branch for feature branches
branches.feature{type}/{ticket}-{description}Branch naming pattern
commits.typesFeature, Fix, Hotfix...Allowed change types
commits.requireTicketfalseEnforce ticket ID
commits.ticketPattern^[A-Z]+-\d+$Regex for ticket validation
issueTracker.typeautolinear, jira, github, none

Error Handling

ScenarioAction
Branch already existsAsk to switch or rename
Ticket required but missingPrompt for ticket ID
Invalid ticket formatShow expected format, ask to retry
Issue tracker API failsWarn and continue without ticket details
Not on expected branchWarn and ask for confirmation
Uncommitted changesWarn and ask to stash or commit first
.pr-context.json existsAsk to overwrite or abort

Examples

Basic Usage (No Config)

User: /start
Agent: What is the ticket ID? -> "PROJ-123"
Agent: What type of change? -> "fix"
Agent: Brief description? -> "fix login timeout"
Result: Branch fix/proj-123-fix-login-timeout created

With Linear Integration (via MCP)

User: /start
Agent: What is the ticket ID? -> "ENG-456"
[Uses Linear MCP to fetch: "Add OAuth2 support"]
Agent: What type? -> "feature"
Agent: Description? -> "oauth2-authentication"
Result: Branch feature/eng-456-oauth2-authentication created
        Ticket: ENG-456 - Add OAuth2 support

Without Ticket (When Not Required)

User: /start
Agent: What is the ticket ID? -> (skip)
Agent: What type? -> "chore"
Agent: Description? -> "update dependencies"
Result: Branch chore/update-dependencies created

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.