expand-tasks

Query Perplexity MCP to research TaskMaster tasks and write findings back to tasks.json.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/samixisme/client-dashboard --skill expand-tasks
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: expand-tasks
Source: https://github.com/samixisme/client-dashboard/tree/main/.agents/skills/prd-taskmaster/companion-skills/expand-tasks
Command: npx skills add https://github.com/samixisme/client-dashboard --skill expand-tasks

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Automates the upfront research phase for TaskMaster tasks by reading tasks.json, launching parallel research agents per task, and writing findings back to the task data. This reduces time to prototype and improves task clarity before implementation.

Core Features & Use Cases

  • Reads tasks.json, generates targeted research prompts for each task
  • Launches parallel research agents in waves with a cap of 5 concurrent agents per wave
  • Collects and writes back research results into tasks.json to guide implementation

Quick Start

Expand all tasks with research after the PRD is parsed into tasks.

Frequently Asked Questions about expand-tasks

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate pre-coding research for TaskMaster tasks?

Automating pre-coding research for TaskMaster tasks involves reading tasks.json, generating prompts, launching parallel Perplexity agents, and writing summarized findings back to guide implementation.

How does parallel research for AI agents work when expanding tasks?

Parallel research for AI agents works by launching concurrent waves capped at 5 agents, querying Perplexity MCP for findings, and collecting status updates for each task.

Can I use Perplexity MCP to research a PRD-derived tasks.json file?

Yes, Perplexity MCP supports researching a PRD-derived tasks.json file through a scripted expansion workflow that generates prompts and writes results back into task entries.

What is the best way to write research findings back to tasks.json?

The best way to write research findings back to tasks.json is using an automated workflow that collects Perplexity MCP results and integrates them into task data for implementation guidance.

Are there concurrency limits when launching parallel research agents in TaskMaster?

Yes, parallel research agents have a concurrency cap of 5 agents per wave to manage status collection and error handling effectively.