agentleFS
renfei-designrenfei-design/Figma-AI-Bridge/.github/skills/online-research/SKILL.md

online-research

Skill16 starsChanged 9 months agoMITView on GitHub

Research any topic using live web sources before creating content or documentation. Use when asked to research, compare, analyze, or create content that requires up-to-date information beyond training data. Fetches authoritative sources, extracts structured data, and produces accurate, sourced content.

1 other project uses a version of this.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 0

    None found.

    Edited · 0

    None found.

    Wrote their own version · 1

    1. sxntixgo/frugal-skills★ 0same name

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • In this repository5 other files it gives its agents.

The file

We split this skill into the rules it gives an agent

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

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Look for dates/timestamps to confirm freshness

    line 68
  2. Distinctive:

    If a page redirects, follow the redirect URL in a second fetch

    line 69
  3. Distinctive:

    If a page returns 403/404, try an alternative source

    line 70
  4. Distinctive:

    After fetching, organize findings:

    line 74
  5. Distinctive:

    Before delivering researched content:

    line 131

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “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

Online Research

Fetch live web data to ground content in current facts. Research first, create second. Use this skill whenever content accuracy depends on information that may have changed since training.

When to Use

  • User asks to "research," "look up," "compare," or "find the latest" on any topic
  • Content involves pricing, specs, rankings, dates, or statistics that change over time
  • User explicitly wants current/accurate data (not training knowledge)
  • Creating a doc or brief on a topic where outdated info would hurt credibility

Prerequisites

  • fetch_webpage tool available (built into the agent)
  • For Figma doc output: combine with the figma-doc skill

Workflow

1. Identify What Needs to Be Current

Before researching, determine:

  • What claims need live sourcing? (pricing, specs, rankings, dates, availability)
  • What can use training knowledge? (concepts, history, frameworks, opinions)
  • What's the user's actual question? (don't over-research tangents)

2. Select Sources

Choose 2–5 authoritative sources per topic. Prioritize:

Source typeWhen to useExamples
Official product pagesPricing, specs, featuresVendor pricing pages, product docs
Aggregator / comparison sitesCross-vendor comparisonsReview sites, benchmark aggregators
Documentation / API docsTechnical specs, API detailsOfficial developer docs
News / blog postsRecent announcements, launchesTech news sites, official blogs
Data sourcesStatistics, market dataIndustry reports, public datasets

Source selection principles:

  • Primary over secondary — official docs over blog summaries
  • Multiple sources for contested claims — cross-reference pricing, benchmarks
  • Recency matters — prefer sources updated within the last 3 months
  • Fallback sources — if a primary source 403s or redirects, use aggregators

3. Fetch and Extract

Use fetch_webpage with:

  • urls: array of 1–3 URLs per fetch call
  • query: specific question to focus extraction
fetch_webpage(
  urls: ["https://example.com/pricing", "https://example.com/specs"],
  query: "current pricing and key specifications"
)

Extraction tips:

  • Tables in fetched content often contain the most useful structured data
  • Look for dates/timestamps to confirm freshness
  • If a page redirects, follow the redirect URL in a second fetch
  • If a page returns 403/404, try an alternative source

4. Synthesize

After fetching, organize findings:

  1. Separate facts from interpretation — facts are sourced, interpretation is yours
  2. Flag stale or conflicting data — if sources disagree, note it
  3. Structure for the output format — if building a doc, organize into sections
  4. Cite sources — include source URLs or names in the content

5. Output

Produce the content in whatever format the user needs:

  • Figma doc → use figma-doc skill with researched content
  • Text response → structured markdown with sourced claims
  • Comparison → tables with sourced data points

Research Patterns

Price Comparison

  1. Fetch each vendor's official pricing page
  2. Fetch one aggregator for cross-reference
  3. Build a comparison table with consistent dimensions
  4. Note pricing date and any caveats (volume discounts, preview pricing)

Product / Feature Comparison

  1. Fetch each product's official specs/docs page
  2. Fetch one independent review or comparison
  3. Identify common dimensions (features, specs, availability)
  4. Build a table or columns comparing across products

Market / Industry Analysis

  1. Fetch 2–3 recent news articles or reports
  2. Fetch official announcements from key players
  3. Synthesize trends, numbers, and quotes
  4. Structure as narrative sections with data tables

Technical Deep-Dive

  1. Fetch official documentation and API references
  2. Fetch community discussions or tutorials for practical context
  3. Structure as overview → details → examples → gotchas

Event / Announcement Summary

  1. Fetch the official announcement or press release
  2. Fetch 1–2 analysis pieces for context
  3. Summarize what happened, why it matters, what's next

Handling Common Issues

IssueSolution
403 / blocked pageTry an aggregator or cached version of the data
RedirectFollow the new URL with another fetch_webpage call
Stale data on pageCheck for a "last updated" date; note it in output
Conflicting sourcesPresent both data points with source attribution
Too much dataFocus extraction with a specific query parameter
Missing dataNote what couldn't be verified; use training knowledge as fallback with disclaimer

Quality Checklist

Before delivering researched content:

  • Every data point that could be outdated is sourced from a live fetch
  • Sources are authoritative (official > aggregator > blog)
  • Conflicting data is flagged, not silently resolved
  • Pricing/dates include an "as of" timestamp
  • Content structure matches the output format
  • Fetched data is synthesized, not just dumped

Notes

  • Don't fetch for everything — conceptual explanations, frameworks, and opinions don't need live data
  • Batch fetches — fetch 2–3 URLs in one call when they're independent
  • Be specific with queries — a focused query extracts better data than a vague one
  • Respect rate limits — don't make more than 5–6 fetch calls per research task
  • Training knowledge is still valuable — use it for context, history, and framing
  • Always disclose — if data couldn't be verified live, say so

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.