anthroclaw
timur-nocodia/anthroclaw
AnthroClaw is a local-first control plane and Pi-native runtime for personal multi-agent assistants operating across channels like Telegram and WhatsApp.
29 6MITUpdated 2026-05-26
73
GoodOverview
AnthroClaw serves as a comprehensive local gateway and agent harness that manages workspace isolation, persistent memory, safety profiles, scheduled tasks, and multi-channel routing. It features a Next.js control UI, a lossless context management plugin, a learning loop for durable skill and memory refinement, and robust support for MCP tools and custom plugins.
Capabilities
- ▸Pi-native runtime execution and harness management
- ▸Multi-channel support for Telegram, WhatsApp, and web chat
- ▸Secure-by-default agent safety profiles
- ▸Lossless context management with hierarchical summarization and FTS5 search
- ▸Automated learning loop to propose memory and skill updates from session transcripts
- ▸Durable scheduled tasks and heartbeat routines
Best for
Running personal multi-agent assistant fleets across Telegram and WhatsApp, Managing agent memories, skills, and safety profiles from a unified control dashboard, Executing automated background routines and scheduled tasks via a local control plane, Automatically converting user corrections and workflow learnings into durable agent skill updates
Works with
Claude CodeCursor
Evoa Score breakdown= Σ (score × weight)
73Task usefulness20%100 → +20.0
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
Code quality15%90 → +13.5
実装・指示の品質。 — AIによるSKILL.md/README解析
Maintenance15%85 → +12.8
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
Documentation12%39 → +4.7
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
Security15%50 → +7.5
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
Originality10%85 → +8.5
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
Popularity8%41 → +3.3
コミュニティの採用度。 — GitHub Stars/Forks（対数スケール）
Compatibility5%50 → +2.5
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
Weighted total72.7 / 100
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
AnthroClaw enforces rigid safety profiles (chat_like_anthroclaw, public, trusted, private) to control tool access and sandbox parameters. The learning system explicitly prevents the storage of secrets, tokens, credentials, and API keys.
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.