ketch
1broseidon/ketch
Ketch is a fast, stateless Go-based command-line interface and Model Context Protocol (MCP) server designed for agentic web search, content scraping, code snippet discovery, and documentation lookup.
Overview
Ketch provides a unified interface for AI agents and developers to perform web operations. It features concurrent batch scraping, multi-engine federated search using Reciprocal Rank Fusion, code grepping, documentation retrieval, and breadth-first web crawling. It operates statelessly with local TTL caching and offers both a CLI and an MCP server integration.
Capabilities
- ▸Web search across multiple backends with federated RRF fusion
- ▸Concurrent URL scraping with smart input detection and automatic JS-shell rendering
- ▸Source code search across platforms like GitHub and Sourcegraph
- ▸Documentation search and resolution
- ▸BFS and sitemap web crawling with depth limits and background processing
- ▸Model Context Protocol (MCP) server support over stdio
Best for
AI agents executing real-time web searches and context scraping during reasoning loops., Developers searching for code snippets across open-source repositories via the command line., Gathering and converting live web pages or sitemaps into clean Markdown for documentation or RAG pipelines.
Works with
実タスクにどれだけ役立つか(機能の豊富さ・用途の明確さ)。 — AIによるcapabilities/use-cases解析
実装・指示の品質。 — AIによるSKILL.md/README解析
リポジトリがどれだけ活発に保守されているか。 — GitHub 最終push日時の新しさ
ドキュメントの充実度・分かりやすさ。 — README/独自要約の情報量
危険・不審な挙動が無いか。 — AIによるセキュリティレビュー
ありふれたラッパーではない独自性。 — AIによる独自性判定
コミュニティの採用度。 — GitHub Stars/Forks(対数スケール)
対応AIエージェントの広さ。 — AIによる対応エージェント判定
ライセンス不明/制限あり(Red)のSkillは総合スコアに0.85倍の補正を適用します。 ランキングはこのScoreのみで決まり、広告で変わりません。 算出方法の詳細 →
Security considerations
The MCP server and CLI execute arbitrary URL fetching and crawling without built-in output filtering or private IP blocking. Running the server grants the connected agent the network posture of the host machine.
Categories
Summary and analysis are original content generated by AI Skills Rank. The skill's source text is not reproduced here — view it on the linked repository.