Install
$ agentstack add skill-luochang212-cc-enhance-web-search-cce ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ● Network access Used
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
> 来自 cc-enhance。-cce 后缀表示此技能由 cc-enhance 项目安装,避免与本地同名技能冲突。
网络搜索
核心原则
- 先读注册表。
~/.claude/tool-registry.json记录了每个工具的实测状态。 - 同层并发。独立通道同时发起,谁先返回用谁。
- 失败即降级,不重试。一个工具失败一次就跳到下一层。
优先级链
Tier 0 — 直接 URL
用户给了 URL 或你知道目标地址 → 不搜索,直接提取页面内容。优先用内置 WebFetch(<1s,覆盖 90% 页面);JS 渲染/反爬页面用 Crawl4AI(需 Python 环境);Crawl4AI 超时或被检测时降级到原生 Playwright 拿 inner_text。
Tier 1 — WebSearch(内置,零配置)
WebSearch 是 Claude Code 内置工具,零依赖、零成本、结构化返回。始终最先尝试。失败(如当前模型不支持)则立即降级到 Tier 2。
Tier 2 — Tavily / SearXNG(并列)
Tavily — ~0.6s 响应,JSON 结构化返回,不限频,英中文均可。1000 次/月。
→ [references/tavily.md](references/tavily.md)
SearXNG — 自部署 Docker,零额度限制,~3s 响应,中文 18 条/Tavily 5 条。
→ [references/searxng.md](references/searxng.md)
Tier 3 — websearchprime MCP
域名过滤、时效性、内容长度控制优于 Tavily。有月额度,备选。
→ [references/web-search-prime.md](references/web-search-prime.md)
Tier 4 — DuckDuckGo(兜底)
纯免费但 ~2s、HTML 需解析、限频。仅在前三层全部不可用时使用,降级前告知用户。
→ [references/duckduckgo.md](references/duckduckgo.md)
降级流程
需要搜索
├─ WebSearch (内置) ──失败──┐
│ ▼
├─ Tavily curl (Bash) ──失败──┐
├─ SearXNG (本地 Docker) ──失败──┤
│ ▼
├─ web_search_prime MCP ──失败──┐
│ ▼
└─ DuckDuckGo curl ── 兜底
搜索拿到 URL 后,提取页面内容:优先 WebFetch(内置,<1s)→ JS 重/反爬页用 Crawl4AI(magic=True 解微信)→ 超时或被检测时降级到原生 Playwright(wait_until=commit)。
结果展示
每条搜索结果必须附带链接,让用户能直接点开看原文。格式:
1. **标题** — 来源
🔗 https://...
维护
工具状态变更时更新 ~/.claude/tool-registry.json。
环境依赖
| 依赖 | 用途 | 安装方式 | 磁盘 | 必需 | |------|------|----------|------|------| | Tavily API Key | 首选搜索,~0.6s,1000次/月 | 注册 tavily.com → export TAVILY_API_KEY=xxx | 0 | 否 | | ddgs | 兜底搜索,防 CAPTCHA | pip install ddgs | ~1MB | 否 |
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: luochang212
- Source: luochang212/cc-enhance
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.