一键在 Manus 中运行任何 Skill

web-research

星标9

分支0

更新时间2026年6月7日 08:13

Research current, external, or web-based information; compare products, pricing, libraries, APIs, companies, docs, releases, issues, or news; and analyze content from URLs.

安装

用 Codex 或 Claude 帮你安装复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。

在 Manus 中运行

来源

panthomakos

panthomakos/dotfiles

打开 GitHub 仓库查看创作者相关仓库

下载

在 Manus 中运行

Web Research

Use this skill when the user asks to research current, external, or web-based information; compare products, pricing, libraries, APIs, companies, docs, releases, issues, or news; or analyze content from one or more URLs.

Workflow

If the user provides specific URLs, use web_fetch on those URLs first.
If the user asks a broad/current question, use web_search to find sources.
Prefer primary and high-quality sources: official docs, changelogs, specs, standards, vendor pages, GitHub repositories/issues, academic or government sources, and reputable publications.
Use web_fetch to inspect promising sources before making detailed claims. Do not rely only on search snippets for important facts.
Cross-check important or surprising claims against multiple sources when practical.
Track dates: distinguish current facts from old posts, archived docs, and stale examples.
In the final answer, include source URLs for important claims and mention uncertainty or source limitations.
Be concise unless the user asks for a deep research report.

Tool Guidance

Use web_search for default discovery. It calls the reusable ~/.pi/agent/bin/pi-web-search CLI: Google curl first, then DuckDuckGo HTML fallback if Google returns JS-only/CAPTCHA/unparseable markup.
Use web_fetch for direct source inspection and analysis. It calls the reusable ~/.pi/agent/bin/pi-web-fetch CLI.
For direct URLs, skip search unless additional context or corroboration is needed.
Use agent-browser via bash only when normal fetch/search is insufficient: JS-heavy pages, clicking through interfaces, screenshots, forms, login/session workflows, or other interactive browser tasks.

同仓库更多 Skills

同仓库

review

panthomakos/dotfiles

Use for strict code-quality review of local changes, commits/ranges, or GitHub PRs; also for review-oriented subagents/checkpoints in orchestrated coding work.

2026-06-239

subagents

panthomakos/dotfiles

Use to delegate context-heavy or independent work to an isolated worker: broad code search, web research, multi-file exploration, long-running implementation, testing/log investigation, review, parallel analysis, or concise summarization of noisy work.

2026-06-239

handoff

panthomakos/dotfiles

Use when compressing context for another session, agent, fork, subagent, queued item, or future continuation. Defines handoff content standards, not file/session mechanics.

2026-06-169

orchestration

panthomakos/dotfiles

Use when coordinating a long-running coding effort through plans, subagents, handoffs, reviews, queues, and checkpoints while keeping the main context small.

2026-06-169

tdd

panthomakos/dotfiles

Test-driven development with red-green-refactor loop. Use by default when implementing features, fixing behavior bugs, changing public behavior, or adding integration tests; also use when user mentions testing, TDD, red-green-refactor, or test-first development.

2026-06-169

grill-me

panthomakos/dotfiles

Interview the user relentlessly about a plan or design until reaching shared understanding, resolving each branch of the decision tree. Use when user wants to stress-test a plan, get grilled on their design, or says "grill me on..." or similar.

2026-06-079

name	web-research
description	Research current, external, or web-based information; compare products, pricing, libraries, APIs, companies, docs, releases, issues, or news; and analyze content from URLs.

Web Research

Workflow

If the user provides specific URLs, use web_fetch on those URLs first.
If the user asks a broad/current question, use web_search to find sources.
Prefer primary and high-quality sources: official docs, changelogs, specs, standards, vendor pages, GitHub repositories/issues, academic or government sources, and reputable publications.
Use web_fetch to inspect promising sources before making detailed claims. Do not rely only on search snippets for important facts.
Cross-check important or surprising claims against multiple sources when practical.
Track dates: distinguish current facts from old posts, archived docs, and stale examples.
In the final answer, include source URLs for important claims and mention uncertainty or source limitations.
Be concise unless the user asks for a deep research report.

Tool Guidance

Use web_search for default discovery. It calls the reusable ~/.pi/agent/bin/pi-web-search CLI: Google curl first, then DuckDuckGo HTML fallback if Google returns JS-only/CAPTCHA/unparseable markup.
Use web_fetch for direct source inspection and analysis. It calls the reusable ~/.pi/agent/bin/pi-web-fetch CLI.
For direct URLs, skip search unless additional context or corroboration is needed.
Use agent-browser via bash only when normal fetch/search is insufficient: JS-heavy pages, clicking through interfaces, screenshots, forms, login/session workflows, or other interactive browser tasks.