edgecrab

Name: edgecrab
Rating: 66 (1 reviews)
Author: raphaelmansuy

MCP Server

raphaelmansuy/edgecrab

EdgeCrab is an advanced Rust-based autonomous agent and development guide featuring a ReAct execution loop, multi-platform gateway, and persistent goal management.

80 17No licenseUpdated 2026-07-24

Good

View on GitHub Open source file

Overview

EdgeCrab is a comprehensive Rust workspace designed for building, testing, and running autonomous AI coding agents. It features a rich crate architecture containing a core ReAct loop, 30+ built-in tools, SQLite-backed session persistence, model routing, a ratatui terminal user interface, and multi-platform messaging gateways. The development guide outlines project layout, crate dependencies, agent builder patterns, persistent goals (the Ralph loop), and system prompt assembly.

Capabilities

▸ReAct tool-call execution loop with context compression
▸Subagent delegation and task execution
▸Persistent agent goals and subgoals surviving conversation resets
▸Multi-platform gateway supporting Telegram, Discord, Slack, and others
▸Path-safe file read, write, patch, and search utilities
▸Sandboxed code execution and headless browser automation

Best for

Developing and customizing autonomous AI coding agents in Rust, Running a terminal-based AI assistant with streaming responses and mission steering, Integrating AI agents across multiple chat platforms like Telegram, Discord, and Slack via gateways, Managing persistent agent memory, tools, and background cron jobs

Works with

Claude CodeCursorGitHub Copilot

Evoa Score breakdown= Σ (score × weight)

Task 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%100 → +15.0

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%48 → +5.8

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%52 → +4.2

コミュニティの採用度。 — GitHub Stars/Forks（対数スケール）

Compatibility5%75 → +3.8

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 total78.2 / 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

Includes security features such as path safety checks, SSRF guards for web requests, and command scanners. External skills and tools should be reviewed for untrusted execution risks.