Skip to content
Back to skills

Using Human Researcher

ASecurity

Help a user get started with Human Researcher, choose the next research task from their current materials, or coordinate several research tasks toward a proposal. Do not use for a clearly scoped paper summary, literature search, idea critique, design, draft, or review that can be handled directly.

  • 3 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 9, 2026
researchgo

Security analysis

A100/100

Pro scans all 2 files and shows the line behind each finding

Scanned October 9, 2026

npx -y skills add taewan2002/human-researcher --skill using-human-researcher --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Using Human Researcher?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Using Human Researcher
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/taewan2002-using-human-researcher/badge)](https://www.skillsdirectory.com/skills/taewan2002-using-human-researcher)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: using-human-researcher
description: Help a user get started with Human Researcher, choose the next research task from their current materials, or coordinate several research tasks toward a proposal. Do not use for a clearly scoped paper summary, literature search, idea critique, design, draft, or review that can be handled directly.
metadata:
  version: 0.1.0
  collection: human-researcher
---

# Using Human Researcher

Honor the user's requested length, language, and format. Fit essential qualifications inside that limit, without an extra preface or closing note. Before sending a count-constrained answer, check the requested number of sentences or items.

Help the researcher choose a useful next action, not a compulsory seven-stage pipeline. Use the current conversation and supplied materials before asking for more information.

## Locate the next decision

Identify the intended deliverable, chosen question (if any), available evidence, constraints, and the uncertainty currently blocking progress. Ask only for missing information that changes the next action. An unfinished study or unknown publication target need not block a discussion draft.

| Current need | Task skill, if available | Useful stopping point |
|---|---|---|
| Understand one supplied paper | `read-paper` | Claim, supporting evidence, and limits understood |
| Find literature or verified citing work | `trace-research` | Relevant evidence and actual coverage recorded |
| Organize collected work or draw a taxonomy | `map-field` | Consistent, evidence-linked classification |
| Choose or sharpen a research question | `develop-idea` | Question, strongest alternative, and first discriminating check |
| Make a selected question testable and feasible | `design-research` | Measures, outcome interpretations, resources, and next decision |
| Write a proposal, section, or discussion draft | `write-proposal` | Requested draft with evidence and unresolved decisions |
| Critique or score an existing draft | `review-proposal` | Located findings and remedies; revision only when requested |

A selected question with competing explanations but no discriminating measure goes to `design-research`; measurement design can refine the test without restarting question selection. For guidance-only requests, describe the next task using existing context instead of turning the recommendation into an intake questionnaire.

## Suggest or carry out the scoped path

Give the shortest useful path, its reason, the needed input, and what will determine the following step. Skip work already supported by evidence and preserve the user's selected direction. A clear drafting or review request goes directly to that task; this guide is not a mandatory entry point.

Check the actual available skill catalog before invoking any task skill. The guide can explain a path on its own; it must not claim another package is installed, invent execution, automatically install dependencies, or require sibling files. If a needed skill is unavailable, explain the task in ordinary language and proceed within the authorized capabilities, or name the missing capability.

When the user asks for guidance only, finish with the recommendation. When they authorize the work, perform the useful next task without adding a routine approval step. Do not recursively invoke this guide or call a sequence of skills an independent review. Separate completed work from proposed next actions, and research planning from experiment execution.

Files in this skill

  • SKILL.md3.4 KB
  • agents/openai.yaml269 B

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.