research

Runs a research-dive workflow producing an approved design doc and discovery report without writing code.

Updated May 27, 2026
One-click install
npx skills add https://github.com/jokerman89/lintel --skill research-jokerman89
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: research
Source: https://github.com/jokerman89/lintel/tree/main/skills/research
Command: npx skills add https://github.com/jokerman89/lintel --skill research-jokerman89

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Teams often need to understand an unfamiliar domain or validate a direction before committing to implementation. This Skill runs a structured explore-and-understand workflow that produces an approved design document and a discovery report, without generating any code. ## Core Features & Use Cases - Composite Research Workflow: Chains the SENSE, DEFINE, and DISCOVER phases while skipping PLAN, BUILD, REVIEW, SHIP, and CAPTURE. - Approved Design Output: Produces an APPROVED design doc plus a discover-report summarizing ADRs, lessons applied, and overlapping skills. - Resumable Session: Findings persist so the operator can later continue into planning with /li:resume if the research validates the direction. - Use Case: A customer asks what a proposed feature would look like. Run the research workflow to survey existing patterns, answer the six DEFINE forcing questions, and deliver an approved design doc before any build commitment. ## Quick Start Ask the agent to run the research skill to investigate a new domain and produce a design doc and discovery report without writing any code.

Frequently Asked Questions about research

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

FAQPage Schema
How do I research a new domain before writing code?▼

Run the research skill, which chains the SENSE, DEFINE, and DISCOVER phases into one workflow. It produces an approved design doc and a discover-report summarizing findings, with no code generated.

What is the difference between research mode and the full cycle?▼

Research mode runs only SENSE, DEFINE, and DISCOVER, skipping PLAN, BUILD, REVIEW, SHIP, and CAPTURE. The full cycle continues through implementation and delivery, so use it when you are ready to build.

Can I continue to implementation after a research session?▼

Yes. The design doc and discover-report persist after the research run, and you can resume with /li:resume to continue into PLAN if the research validates the direction. Findings can also be captured standalone with /li:capture.

When should I not use the research workflow?▼

Skip it when you are ready to build, when the territory is already known, or when you only need a quick lookup on a single topic. In those cases use the full cycle, /li:define, or a focused web search instead.

How much time and tokens does a research run cost?▼

A typical research-dive run costs roughly 10-20k tokens and takes 20-40 minutes. It is a read-only workflow with no compliance gates, so it suits pre-commit exploration.