Deep Research Skills
Deep Research Skill
A structured deep-research skill for Claude Code, OpenCode and Codex with human control at every step
- 2,438 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- The skill launches background web agents and writes report files
Install
- Works in
Manual install
git clone https://github.com/Weizhena/deep-research-skills.git
cp -r deep-research-skills/skills/research-en/* ~/.claude/skills/
cp deep-research-skills/agents/web-search-agent.md ~/.claude/agents/
pip install pyyamlFor the Chinese version use research-zh.
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The skill launches background web agents and writes report files
This is third-party code. Review the repository files before installing.
What it does
The skill runs research in two phases with a human in the loop at each stage. First it builds a plan from the topic: a list of objects to study and a set of fields to collect, with confirmation and the option to add items and fields. Then it runs deep investigation, where a background agent searches the web and gathers data for each object. Finally a separate command assembles a markdown report with a table of contents. There are English and Chinese versions, plus variants for Claude Code, OpenCode and Codex. Web search in OpenCode needs an environment variable, and the skill itself requires the pyyaml dependency.
Who it is for. For analysts, researchers and product managers who need a controlled topic survey with sources.
Good fit when
- You prepare a survey of technologies, tools or competitors
- You need a structured report with fields per object
- You want to keep control over the plan and coverage
Not a fit when
- You need a quick one-off answer without a plan or report
- You have no web search access in your environment
Example request
Build a research outline on AI agents 2025 and gather data for eachLimitations
The skill requires the pyyaml dependency and a web agent installed, and in OpenCode needs an environment variable for full search. Report quality depends on web access and on the user's confirmations at each step.
How to disable. Remove the research skill folders from your agent's skills directory and the linked web agent from the agents directory.
FAQ
Which agents are supported?
Claude Code, OpenCode and Codex, with separate skill and web-agent versions for each.
What is needed for web search?
The pyyaml dependency, and in OpenCode also an environment variable to enable extended search.
Discussion
No questions or reviews yet.
Related
SkillsA set of 13 SKILL.md skills: research, a product doc, decks, articles and audits for any compatible agent
Skills that turn a coding agent's output into plain language: honest reports, straight verdicts, followable steps
A skill that runs user research inside the agent: plan, interview guide, report synthesis, plus interviews and quantitative surveys through the Cookiy AI servic
A skill that builds an anonymized developer profile and a vertical poster from local Claude Code and other CLI history
Seven Claude Code skills: turn a video, article or PDF link into text and an action plan, plus a Scrum coach, a session log and an unblock helper
A library of 229 skills for work and life: sales, marketing, finance, engineering, design and parallel agent runs
README in short
The README describes a structured two-phase research workflow: outline generation and deep investigation, with a human in the loop at each stage. Installation is by cloning, copying the right skill version and web agent, and installing the pyyaml dependency, with an environment variable needed for OpenCode. The research, research-add-items, research-add-fields, research-deep and research-report commands move from topic to report. English and Chinese versions exist. MIT licensed.
README badge
Original SKILL.md
--- name: research user-invocable: true allowed-tools: Read, Write, Glob, WebSearch, Task, AskUserQuestion description: Conduct preliminary research on a topic and generate research outline. For academic research, benchmark research, technology selection, etc. --- # Research Skill - Preliminary Research ## Trigger `/research <topic>` ## Workflow ### Step 1: Generate Initial Framework from Model Knowledge Based on topic, use model's existing knowledge to generate: - Main research objects/items list in this domain - Suggested research field framework