agentleFS
joinwell52-AIjoinwell52-AI/CodeFlowMu-open/skills/web-research/SKILL.md

web-research

Skill2 starsChanged 4 months agoMITView on GitHub

Orchestrate web search and web extraction to produce evidence-based conclusions or reports with traceable source URLs. Use for web research, market or competitor scans, current-fact verification, technical investigation, policy research, multi-source comparison, dynamic-page research, table-backed analysis, or any answer that must cite web sources.

45 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use data-analysis, meeting-notes and 9 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • shanggqmSkill

    data-analysis

    数据分析技能,提供结构化的数据查询、统计分析和报告生成工作流。当用户需要查询数据库、进行数据统计分析、生成分析报告时使用。按规范输出原始数据、计算结果、Markdown报告和HTML可视化报告。

    Also in 5 projects that use this

  • claude-office-skillsSkill

    meeting-notes

    >

    Also in 4 projects that use this

  • kongyo2Skill

    prompt-engineering

    Comprehensive prompt engineering framework for designing, optimizing, and iterating LLM prompts. This skill should be used when users request prompt creation, optimization, or improvement for any LLM task, or when users need help translating vague requirements into effective prompts through collaborative dialogue and iterative refinement.

    Also in 4 projects that use this

  • rogue-dev-studioSkill

    browser-automation

    Canonical browser automation and web UI verification: Playwright, agent browser CLIs, and CDP-based tooling in one playbook.

    Also in 3 projects that use this

  • JetBrainsSkill

    brand-guidelines

    Applies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply.

    Also in 4 projects that use this

  • AnthropicSkill

    pdf

    Comprehensive PDF manipulation toolkit for extracting text and tables, creating new PDFs, merging/splitting documents, and handling forms. When Claude needs to fill in a PDF form or programmatically process, generate, or analyze PDF documents at scale.

    Also in 6 projects that use this

The file

Show its rules and text

We split this skill into the rules it gives an agent

14 rules. 0 are common, 9 are distinctive, and 5 are very specific to this project.

  • 0

    Common

    Most of the 38 edited versions we read keep these.

  • 9

    Distinctive

    This author's own choices. Worth a look.

  • 5

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Run an end-to-end research workflow by combining web-search and web-extract.

    line 8
  2. Very specific:

    Use the web-search procedure to construct queries and select authoritative, diverse sources.

    line 13
  3. Very specific:

    Use the web-extract procedure on the selected URLs:

    line 14
  4. Distinctive:

    Use Playwright only for dynamic content when needed.

    line 16
  5. Distinctive:

    Write the requested answer or report with findings, evidence, limitations, and source list.

    line 24
  6. Distinctive:

    When operating inside an agent task, use the available report tool and include source URLs in the report body or structured evidence.

    line 25
  7. Very specific:

    Do not generate a site-specific .mjs probe for normal work.

    line 26
  8. Distinctive:

    When the research hierarchy is known, execute it in dependency order.

    line 27
  9. Distinctive:

    Every material factual claim must be supported by an extracted source.

    line 57
  10. Distinctive:

    Use exact dates when freshness affects the conclusion.

    line 58
  11. Distinctive:

    Prefer primary and official sources.

    line 59
  12. Distinctive:

    Do not cite a search result that was never opened and checked.

    line 61
  13. Distinctive:

    Do not cite a URL that does not support the associated claim.

    line 62
  14. Very specific:

    Every evidence item retained source_url.

    line 70

“Common” is measured: kept by at least 6 in 10 of the 38 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Web Research

Run an end-to-end research workflow by combining web-search and web-extract. Produce conclusions that remain traceable to the underlying webpages.

Workflow

  1. Define the research question, decision to support, scope, freshness requirement, and expected report shape.
  2. Use the web-search procedure to construct queries and select authoritative, diverse sources.
  3. Use the web-extract procedure on the selected URLs:
    • Open webpages.
    • Use Playwright only for dynamic content when needed.
    • Extract relevant正文.
    • Extract relevant tables with headers and units.
    • Preserve source_url on every evidence item.
    • Set explicit extraction quality gates; reject weak pages instead of treating navigation text as evidence.
  4. Compare sources. Resolve differences using source authority, dates, definitions, methods, and directness of evidence.
  5. Separate verified facts, reasonable inferences, conflicting evidence, and unknowns.
  6. Form conclusions only from extracted evidence. Attach one or more source URLs to each material claim.
  7. Write the requested answer or report with findings, evidence, limitations, and source list.
  8. When operating inside an agent task, use the available report tool and include source URLs in the report body or structured evidence.
  9. Do not generate a site-specific .mjs probe for normal work; configure the shared web tools instead.
  10. When the research hierarchy is known, execute it in dependency order. Complete the inventory layer before sampling descendants, and continue automatically while each next step is read-only and in scope.

Research Report Contract

## Research Question
<question and scope>

## Method
<queries, source-selection rule, retrieval date, dynamic-page use>

## Findings
1. <finding> [source](source_url)
2. <finding> [source](source_url)

## Table Evidence
<normalized table or comparison, with source_url for each source/table>

## Conclusion
<answer, recommendation, or decision support>

## Uncertainty And Conflicts
<unknowns, stale data, conflicting definitions, inaccessible sources>

## Sources
- <title> - <source_url>

Source Discipline

  • Every material factual claim must be supported by an extracted source.
  • Use exact dates when freshness affects the conclusion.
  • Prefer primary and official sources; use secondary sources for context or comparison.
  • Clearly label inference rather than presenting it as a sourced fact.
  • Do not cite a search result that was never opened and checked.
  • Do not cite a URL that does not support the associated claim.

Completion Criteria

  • Searches were purposeful and recorded.
  • Selected pages were opened and inspected.
  • Dynamic content used Playwright only where necessary.
  • Relevant正文 and tables were extracted.
  • Every evidence item retained source_url.
  • Conclusions distinguish fact, inference, conflict, and uncertainty.
  • The final answer or report includes usable source links.
  • The workflow did not stop at a proposed method when the next research action was already authorized and executable.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.