Hand-crafted Claude Code skills for higher-quality agent results.
A curated set of context-engineering skills: when to compress, when to summarize, when to spawn a subagent, how to checkpoint long sessions. Built and battle-tested by NeoLab. Compatible with OpenCode, Cursor, Antigravity, Gemini CLI.
Добавьте в context_servers. Zed перезагружается автоматически.
claude mcp add context-engineering-kit-skill -- git clone https://github.com/NeoLabHQ/context-engineering-kit ~/.claude/skills/context-engineering-kit
Однострочная команда. Проверить: claude mcp list. Удалить: claude mcp remove.
Сценарии использования
Реальные сценарии: Context Engineering Kit
Keep a multi-hour coding session from blowing the context window
👤 Engineers on long agent sessions⏱ ~15 minintermediate
Когда использовать: You're 3 hours into a coding loop and the model is starting to drop context.
Предварительные требования
Server/skill installed and authenticated — See repo README
Поток
Checkpoint
Use the context-engineering skill's 'checkpoint' to summarize what we've done into a tight handoff doc.✓ Скопировано
→ A handoff doc Claude can re-read instead of full history
Resume
Compact the conversation and resume from the handoff doc.✓ Скопировано
→ Session continues with fresh context
Итог: Sessions that don't degrade as they grow.
Подводные камни
Checkpointing too often loses fine-grained context. The kit suggests every ~30 turns or before any /compact. — Checkpointing too often loses fine-grained context. The kit suggests every ~30 turns or before any /compact.
Стоимость и лимиты
Во что обходится
Квота API
See provider docs for rate limits
Токенов на вызов
Varies by tool
Деньги
See repo README for pricing details
Совет
Cache tool results and avoid repeated identical calls.
Безопасность
Права, секреты, радиус поражения
Хранение учётных данных: Use environment variables; never commit secrets
Исходящий трафик: Tool calls go to the provider's API as documented
Устранение неполадок
Частые ошибки и исправления
checkpoint doc too long
Tune the skill's max-length; default is ~1500 tokens. Anything bigger and the doc itself eats context.
Проверить: wc -w on the checkpoint
subagent doesn't have enough context
The skill's spawn template requires a 'mission' block — fill it explicitly, don't let it default.