research

Investigates questions against primary sources and saves cited findings as Markdown files.

Updated Aug 19, 2026
One-click install
npx skills add https://github.com/PVMalove/claude-agent-harness --skill research-pvmalove
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: research
Source: https://github.com/PVMalove/claude-agent-harness/tree/main/skills/vendor/mattpocock/engineering/research
Command: npx skills add https://github.com/PVMalove/claude-agent-harness --skill research-pvmalove

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Getting accurate answers about APIs, libraries, or technical topics usually means interrupting your own work to read documentation and source code. This Skill delegates that reading legwork to a background agent so you keep working while the research happens in parallel. ## Core Features & Use Cases - Primary-source investigation: Follows every claim back to official docs, source code, specs, or first-party APIs instead of relying on secondary write-ups. - Cited Markdown reports: Writes findings to a single Markdown file with a source citation for each claim. - Convention-aware placement: Saves the report where the repo already keeps such notes, matching existing conventions or choosing a sensible location. - Use Case: You need to know how a third-party API handles rate limiting before writing integration code. Launch the research agent, continue coding, and later read a cited Markdown summary of the official rate-limit documentation. ## Quick Start Ask the agent to research how a specific library or API feature works using official documentation and save the findings as a Markdown file in the repo.

Frequently Asked Questions about research

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

FAQPage Schema
How do I delegate technical research to a background agent?▼

Invoke the research skill with your question, and it spins up a background agent that reads primary sources while you continue working. The agent writes its findings to a cited Markdown file in the repository.

What sources does the research agent use for answers?▼

The agent investigates only primary sources: official documentation, source code, specifications, and first-party APIs. It follows every claim back to the source that owns it rather than relying on blog posts or secondary summaries.

Where does the research report get saved?▼

The report is saved as a single Markdown file where the repository already keeps similar notes, matching existing conventions. If no convention exists, the agent picks a sensible location and tells you where it placed the file.

Does the research output include citations for each claim?▼

Yes, each claim in the Markdown report is cited to its source. This lets you verify facts directly against the official documentation or code rather than trusting an unsourced summary.

When should I not use a background research agent?▼

Avoid it for questions answerable from code already in your local context or when you need an immediate blocking answer. The background approach suits open-ended investigation where parallel work is more valuable than instant results.