Skip to content
SkillModel-led1,040 tokens in use

deep-research

A bootstrap skill that sets up the hyperresearch CLI in your project and hands deep research requests to its 16-step pipeline, for Claude Code and OpenAI Codex.

4,944 stars on the repository MIT Licensejordan-gibbs/hyperresearch · skills/deep-research
npx skills add jordan-gibbs/hyperresearch -a claude-code
View source

Install command as written in the repository's README

Skill label

measured from its files

What installing deep-research costs

104tokens idle
1,040tokens in use

Idle is what sits in context while the skill is installed. In use is its instructions, loaded when it runs. Counts are approximate, made offline with cl100k_base.

Model dependence Model-led

Script-ledMixedModel-led

Mostly instructions, so how well it works depends on the model that follows them.

This describes the skill's structure, how much of it is scripts. It does not say how well it works.

Touches

shelldetectednetworkdetected

Declared tags come from the skill's frontmatter; detected ones from its scripts, its code blocks and the commands it names in inline code. This is detected, not audited.

How it scores

Quality

How well the skill is written, counted from its own files: its frontmatter, its steps, examples, stated output and constraints, its size and whether every file it ships is one it names.How we score it →
Good
  • Frontmatter15/15
  • Structure15/25
  • Examples12/15
  • Output5/15
  • Size15/15
  • Resources0/15

Security

Starts at 100 and loses points for each thing in the skill's files that a reader should look at first.How we score it →
Excellent
  • Runs shell commands−5

    does this on the machine, and the frontmatter does not say so

    SKILL.md, line 21

Repository health

Pre-computed score based on development activity, maintenance, community, maturity, and trend momentum.How we score it →
Excellent
  • Development activity96
  • Maintenance100
  • Community64
  • Maturity28
  • Momentum40

deep-research is the entry point to hyperresearch, a deep research harness. When you ask for a research report, a literature review or a multi-source analysis, the skill checks that the hyperresearch command is installed, sets up the project for the agent you are running in, and passes your request on word for word.

The research itself runs in the pipeline the install adds: a tier-adaptive sequence of steps with subagents, a persistent source vault and citation checks before a report ships. Quick lookups that one or two searches can answer are out of scope.

What You Get

  • A check that the hyperresearch CLI is present, with a clear message if it is not
  • A project install that adds the vault, the entry skill, step skills, subagents and a guard hook
  • A verbatim hand-off of your research request to the full pipeline
  • Separate setup branches for Claude Code and OpenAI Codex, including the sandbox flags Codex needs

Common Use Cases

  • Asking for a cited research report on a market, a technology or a policy question
  • Running a literature review that draws on scholarly sources and open-access full text
  • Comparing products or vendors from filings and primary documents with every claim cited
  • Continuing research in a project whose vault already holds sources from earlier runs

What Makes It Unique

Most research skills are a single prompt. This one is a thin installer for a full harness: a 16-step pipeline loaded one step at a time, adversarial critics, a cite-checker and a ship gate, plus a vault that keeps every source for the next session. The skill itself stays small and sends the agent straight to the installed pipeline.

Under The Hood

The skill runs shell commands: it calls the hyperresearch command to check its version and, when the project lacks the entry skill, runs the hyperresearch install command, which writes files into the project (a vault folder, a research folder, a block in CLAUDE.md or AGENTS.md, skills, subagents and a hook under .claude or .codex). It installs the Python package only if you say so. The research that follows fetches web pages and scholarly sources over the network and writes notes to disk, and under Codex it needs a writable, networked sandbox.

Join founders building with open source

Opinionated takes, migration guides, cost-saving tips, and insights from the open source ecosystem.

Subscribe on SubstackJoin 750+ subscribers