louchi1984-coder/deepseek-claude-code-worker-mcp
DeepSeek V4 code worker MCP for Codex Desktop, powered by Claude Code
SUMMARY AI要約 by gpt-5-mini
DeepSeek Code Worker MCP is a background "worker" server for Codex Desktop that delegates heavy code reading, editing, and checking to DeepSeek V4 via the Claude Code CLI. It’s intended for Codex Desktop users who want to offload expensive main-thread token usage to a separate model/worker (saving ~40–60% Codex tokens on suitable tasks). Key features: async job tools (start/get/tail/wait/cancel), synchronous tiny-edit tool, long-running DeepSeek support, scoped Claude Code permissions, setup/doctor checks, and a small claude-deepseek launcher. Outputs compact job status, changed files, snapshot diffs, checks, policy info, and optional detailed logs/events. Requirements: Node.js 20+, Claude Code CLI, DeepSeek API key (macOS/Linux primary). Includes model-selection presets and guidance to minimize Codex host token use. Not a standalone DeepSeek client; used as an MCP server for task delegation.
DETECTED 検出されたAIスタック
このRepoのdescription / GitHub topics / READMEのAI要約に出現したAI関連キーワードをカテゴリ別に表示。各バッジは該当カテゴリの詳細ランキングへリンクします。
使用言語(バイト数比)
オーナー情報
日付
| GitHub作成日 | 2026-05-07 |
| 最終Push | 2026-05-09 |
| 当サイト初検出 | 2026-05-09 |
| 最終取得 | 2026-05-09 14:18 |
類似Repo (同じ言語のAI関連Repo)
Box — self-hosted Claude Code in your pocket. Run real Claude Code & Codex sessions on your own server, driven from your phone (native chat UI, voice, files, skills, always-on).
mohamedzhioua/proofguardKill-tested guard skills for AI coding agents , self-invoking quality gates that catch AI failure modes in security, tests, docs, dependencies, and diffs before the agent says "done." Works with Claude Code, Codex, and Cursor.
IFTC-XLKJ/aiproxy renezander030/blender-cliDrive Blender headless from your AI agent: scaffold, build, animate from natural language. JSON in, JSON out.
muhammad1-wq/mechapulseAI ROBOTICS COMPANY
aditya89bh/ai-guardrails-explanations-design-systemA documentation-first design system for enterprise AI teams building trust and safety into AI products — patterns for warnings, explanations, permissions, uncertainty, refusal, escalation, and recovery.
the-snowmen/fresh-eyesA cast of fictional external users who review your running app in a real browser — isolated from your code, so the feedback stays honest. A Claude Code plugin.
pratik-chiwande/E-commerceI made Ecommerce site using tailwind-css ,React, also used Claude and Chatgpt