github-task-multiple.template

Codex Runtime Compatibility

This prompt was authored for Claude-style slash workflows. In Codex runtime, adapt tool calls as follows:

• AskUserQuestion => ask the user directly in chat.
• WebSearch/WebFetch => use available web tools (search_query, open, find) and cite links.
• Read/Write => use shell/filesystem tools in this workspace.
• Claude-only MCP calls (for example Playwright MCP names) => use available equivalents or clearly state fallback.
• Keep intent and output format identical; only adapt execution mechanics.

---

description: Batch GitHub task operations model: sonnet

/github-task-multiple

Create multiple GitHub issues from a complex prompt by intelligently breaking it into distinct tasks and creating issues in parallel.

Usage

/github-task-multiple <description>
/github-task-multiple "Add auth with OAuth, JWT tokens, password reset, and 2FA"

Features

• Intelligent prompt analysis and task decomposition
• Parallel agent execution using Sonnet 4.5
• Post-creation integration with Phase Dev, Roadmaps, Epics
• User confirmation before issue creation

Instructions

Step 1: Analyze the Prompt

Parse the user's description to identify distinct tasks:

Detection Signals:

• Conjunctions: "and", "plus", "also", "additionally", "as well as"
• Comma-separated lists: "X, Y, and Z"
• Numbered items: "1. First 2. Second"
• Multiple action verbs: "add", "create", "implement", "fix", "update"
• Multiple feature keywords: "feature", "component", "system", "service"

Analysis Algorithm:

function analyzePrompt(prompt) {
  const indicators = {
    conjunctions: (prompt.match(/\b(and|plus|also|additionally|as well as)\b/gi) || []).length,
    listItems: (prompt.match(/,\s*(?:and\s+)?[a-z]/gi) || []).length,
    numbered: (prompt.match(/\d+\.\s+\w/g) || []).length,
    actions: (prompt.match(/\b(add|create|build|implement|fix|refactor|update|integrate)\b/gi) || []).length,
    features: (prompt.match(/\b(feature|component|system|service|module|endpoint|api)\b/gi) || []).length
  };

  const score = indicators.conjunctions * 2 +
                indicators.listItems * 1.5 +
                indicators.numbered * 3 +
                indicators.actions * 1 +
                indicators.features * 0.5;

  return {
    shouldSplit: score >= 4,
    estimatedTasks: Math.max(2, Math.min(8, Math.ceil(score / 2)))
  };
}