Branch and task management

Skill synaptiai/synapti-marketplace/plugins/flow/skills/branch-and-task-management

The Synapti Marketplace is a curated collection of Claude Code plugins designed for AI-augmented development + advanced analytical and research tasks. Each plugin provides specialized agents, skills, and commands that extend Claude Code's capabilities in specific domains.

Install

npx -y skills add synaptiai/synapti-marketplace --skill branch-and-task-management

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

6 stars6 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Create feature branches with naming conventions, load full issue context and impact analysis, and decompose acceptance criteria into atomic parallel tasks with dependencies. Use when starting work on a GitHub issue. This skill MUST be consulted because starting code without context causes misaligned implementations and wasted effort.

SKILL.md

4.0 KB, as published. Nobody here has run it

Branch and Task Management

Domain skill for starting work: branch setup, context loading, and task decomposition.

Iron Law

NO CODE BEFORE CONTEXT. Read the issue, load the history, understand the scope — then create the branch.

Starting a branch without loading issue context leads to misaligned implementations and wasted effort.

Pre-Conditions

Before creating a branch, confirm:

Issue exists and is open (not already closed/resolved)
Issue has acceptance criteria (if not, ask the user or create them)
No existing branch already addresses this issue
You've read the full issue body AND comments (not just the title)

Branch Creation

Follow project conventions from settings or CLAUDE.md:

DEFAULT_BRANCH=$(gh repo view --json defaultBranchRef --jq '.defaultBranchRef.name' 2>/dev/null || echo "main")
git fetch origin "$DEFAULT_BRANCH"
git checkout -b "feature/issue-{N}-{desc}" "origin/$DEFAULT_BRANCH"

Branch naming patterns (from settings.json → conventions.branchPatterns):

feature/issue-{N}-{desc} — New features
fix/issue-{N}-{desc} — Bug fixes
docs/issue-{N}-{desc} — Documentation

Keep {desc} to 3-5 words, kebab-case, meaningful.

Issue Context Loading

Fetch issue details in parallel:

# Parallel: issue details + comments + linked issues
gh issue view $N --json title,body,labels,assignees,milestone
gh issue view $N --comments

Extract from issue body:

Title: One-line summary
Acceptance criteria: Each - [ ] item becomes a task
Labels: Inform implementation approach
Related issues: Cross-references for context

Impact Analysis

Before implementation, identify affected areas:

# Search for related code
grep -r "keyword_from_issue" --include="*.{rb,js,ts,py}" -l
# Check recent changes in related areas
git log --oneline -10 -- "path/to/related/"

Map acceptance criteria to likely file changes. Flag if the issue touches:

Multiple modules (coordination needed)
Shared utilities (risk of side effects)
Test fixtures (may need updates across suites)

Task Decomposition

Convert acceptance criteria to tasks using TaskCreate:

For each acceptance criterion:
  TaskCreate(
    subject: "Implement: {criterion summary}",
    description: "Acceptance criterion: {full text}\nLikely files: {paths}\nVerification: {how to check}"
  )

Rules:

One task per acceptance criterion (minimum)
Add infrastructure tasks if needed (migrations, config)
Add a verification task at the end
Set dependencies with addBlockedBy for sequential work

Parallel Task Detection

Identify tasks that can run concurrently:

Tasks touching different files with no shared imports → parallelizable
Tasks in different directories → likely parallelizable
Tasks modifying the same file → sequential (dependency)

If agent teams are enabled and >5 acceptance criteria with independent file sets → suggest team dispatch.

Decision Journal Init

Create {JOURNAL_DIR}/issue-{N}.md (journal dir defaults to .decisions/):

# Decision Journal: Issue #{N} — {title}
**Issue**: #{N} | **Branch**: {branch-name} | **Started**: {YYYY-MM-DD}
---

The autonomous-workflow skill governs ongoing journal entry format (Init/Log/Summarize).

Verification

Branch setup is valid when:

Branch name matches convention pattern
Issue details loaded successfully
At least one task created per acceptance criterion
Dependencies set correctly (no circular deps)
Journal initialized