跳到正文
cccskills
免费在 GitHub 公开

web-research

Evidence-driven web and open-source research for questions that require multiple sources, factual verification, comparisons, or an adopt/adapt/build decision. Use for substantial research in foreground sessions, the genesis-researcher subagent, and research-profile background sessions. Skip for a single stable fact or a known URL that only needs fetching.

中文简介正在准备中。当前显示原文说明。

查看安装方法

包含的文件(1)

  • SKILL.md4.5 KB

SKILL.md(原文)

安装前可以查看交给 AI 代理的指令内容。

Web Research

Frame the decision

State the question, the decision it informs, the required freshness, and what would count as enough evidence. Choose the lightest depth that can answer it:

  • Focused: one narrow claim or known source. Verify directly and stop.
  • Substantial: several claims, unfamiliar subject matter, comparison, or recommendation. Decompose into independent questions and synthesize them.
  • Exhaustive: high-consequence or broad landscape work. Add coverage tracking, competing explanations, and a final gap pass.

Depth follows the task. Do not add an extra pass merely because the word “research” appears.

Gather evidence

  1. Decompose substantial work into questions whose results can be checked independently. Parallelize only independent searches; follow-ups that depend on an earlier result stay sequential.
  2. Search with web_search, then fetch the sources that carry the actual claim with web_fetch. Use browser tools only when interaction or rendered state is necessary.
  3. For software capability questions, search for existing implementations before proposing a build. Use recon_github_search for public repositories and issues, then recon_github_read to inspect repository metadata, trees, and source. Repository search accepts GitHub query syntax; issue search treats query as literal text and exposes repository, state, and labels as structured filters. GitHub's code-search API requires credentials, so public-only research must discover candidate repositories first and inspect their trees/files rather than silently inheriting operator access. When the capability could run inside a Claude Code session (a hook or guard, status line, prompt or context injection, tool interception, notifications, an in-terminal pane), also search Claude Code mods and plugins: the mod-finder plugin's search_mods tool if installed, otherwise the catalogue at https://raw.githubusercontent.com/karanb192/awesome-claude-code-mods/main/catalogue.md, plus Anthropic's directory (https://raw.githubusercontent.com/anthropics/claude-plugins-official/main/.claude-plugin/marketplace.json, reachable with web_fetch from any session). Record each candidate's access footprint and the CC version it validates on; mods run unsandboxed. A background session cannot search the community catalogue yet (#3321): report that gap rather than "no mod exists". When the foreground /evaluate command is available, apply it for the final adopt/adapt/build comparison. In background sessions, perform that comparison directly: assess how each candidate helps, does not help, could help through adaptation, and what Genesis can learn from it; then classify it as adopt, adapt, learn from, or build.
  4. Prefer primary sources for product behavior, specifications, code, pricing, and current policy. Use strong secondary sources for context or independent testing. A primary source can establish its own documented behavior; independent corroboration matters when incentives, interpretation, or real-world performance are disputed.
  5. Record the URL, publication or update date when relevant, the exact claim supported, and limits or conflicts. A search snippet is a lead, not evidence.

Evidence standard

  • Separate observed facts, source claims, and inference.
  • Resolve conflicts by checking recency, authority, methodology, and scope. Report material conflicts that remain.
  • Treat a failed search or tool call as unknown. Never turn it into “nothing exists.”
  • For recommendations, compare candidates against the stated requirements, including maintenance, license, integration cost, and reversibility.
  • Calibrate confidence per major conclusion: high when evidence is direct and current, medium when evidence is indirect or incomplete, low when key uncertainty remains.

Deliver

Lead with the answer and the decision-relevant evidence. Cite every externally checkable claim near the claim. Include the searches or coverage boundaries only when they explain confidence or remaining uncertainty. End with concrete next actions and open questions when either exists.

For autonomous blocker research, preserve the caller's required JSON result shape after applying this method.

评价

还没有评价。欢迎分享你的使用感受。

评论

无需登录即可发表。欢迎分享使用感受、提问,或说明在哪些环境下无法运行。

还没有评论。来写第一条吧。

提交后会立即公开。请勿填写邮箱、电话等个人信息。URL 最多 2 个。

0 / 1000 字

此表单受 reCAPTCHA 保护,适用 Google 的隐私权政策和服务条款。

同一仓库的技能

Forward Deployed Engineer delivery contract for AWS engagements, build-first artifacts, grounded cost estimates, Well-Architected review, evolution roadmap

中文简介正在准备中。当前显示原文说明。

WingedGuardian/GENesis-AGI992026年10月22日 更新

Canonical guide to Genesis browser automation - layers (Camoufox, Chromium, the user's Chrome over CDP, TinyFish, desktop), per-tool timeouts, safety gates, verify-after-act, what a click checks (scroll, hit test, covered targets), overlays, iframes, tabs, and failure diagnosis

中文简介正在准备中。当前显示原文说明。

WingedGuardian/GENesis-AGI992026年10月22日 更新

This skill should be used when a session BUILDS one Genesis item: implementing a change in a worktree through to opening its PR, including rework from a send-back spec. Load `genesis-development` (the shared core) first. Do NOT load it for driving open PRs to merge (`closing-session`), post-merge verification (`validating-merges`), or specifying a change before it is planned (`spec-session`).

中文简介正在准备中。当前显示原文说明。

WingedGuardian/GENesis-AGI992026年10月22日 更新

cc-update

免费

Update Claude Code (the CC CLI / "clog code") to a new version, or bump the pinned CC version. Use when the user asks to update Claude Code, bump the CC pin, evaluate a new CC release, or says "clog code update". Routes to the canonical, standardized process in docs/reference/cc-compatibility.md — do NOT re-derive the update mechanism by grepping every time. Do NOT use for general "what changed in CC" trivia with no intent to update.

中文简介正在准备中。当前显示原文说明。

WingedGuardian/GENesis-AGI992026年10月22日 更新

This skill should be used when a session's job is to DRIVE OPEN PRs TO MERGE rather than to write new code — "close out the open PRs", "review and fix the open PRs", "what's blocking our PRs", "which PRs are mergeable". It owns the In Review column: it reads each PR's gate status, verifies and fixes review findings on PRs OTHER sessions built, replies in-thread, and merges a PR only under the owner's authorization: a per-PR yes, or the owner's request for closing work covering a PR whose every gate passes. Do NOT load it for building a feature and opening its PR — that is a build session (`build-session`).

中文简介正在准备中。当前显示原文说明。

WingedGuardian/GENesis-AGI992026年10月22日 更新

Code understanding tool selection. Use when exploring architecture, finding definitions, tracing call chains, assessing blast radius of changes, or debugging code paths in the Genesis codebase.

中文简介正在准备中。当前显示原文说明。

WingedGuardian/GENesis-AGI992026年10月22日 更新

查看 WingedGuardian 的全部技能

报告此技能的问题