ascend-ai-coding/awesome-ascend-skills/long-task
Claude Code and agent skill by ascend-ai-coding/awesome-ascend-skills
Trust tier 3
- community
Source check pending
Checking that the SKILL.md file is still in the repository.
长耗时任务管理,支持checkpoint记忆、agent休息与恢复
Install
npx vskill@latest install ascend-ai-coding/awesome-ascend-skills/long-taskSKILL.md
ssh-long-task
概述
管理预计耗时>2分钟的远程任务,通过checkpoint机制支持agent休息和跨会话恢复。
适用场景
- 模型训练(>10分钟)
- 大规模数据处理
- 编译构建任务
- 数据库迁移
流程步骤
1. 启动任务并写入checkpoint
使用connect/scripts/ssh-job.sh start启动任务,获取job_id后立即写入checkpoint:
产出: checkpoint文件 ~/.ssh-jobs/<job_id>/checkpoint.md
2. 告知用户并休息
Agent输出:
3. 恢复时读取checkpoint
新会话或用户请求检查时:
产出: 解析checkpoint内容(task, started, duration, next_steps, context)
4. 智能状态检查
根据预计耗时决定检查策略:
- <5分钟: 每30秒检查一次
- 5-30分钟: 每2分钟检查一次
- >30分钟: 每10分钟检查一次,或用户手动触发
5. 输出检查(Token优化)
先查大小:
按大小决策:
- <10KB: 读取全部
bash connect/scripts/ssh-job.sh output <profile> <job_id> - 10KB-100KB: 只读尾部
bash connect/scripts/ssh-job.sh output <profile> <job_id> --tail 50 - >100KB: 先搜索错误
bash connect/scripts/ssh-job.sh output <profile> <job_id> --grep 'error|fail|exception'
6. 执行next_steps
任务完成后,按checkpoint中的next_steps继续: 逐步执行,每步完成后更新checkpoint或删除。
Checkpoint格式
存储在 ~/.ssh-jobs/<job_id>/checkpoint.md:
Token约束
- 输出检查:先查大小,按大小分级读取(全部/尾部/grep)
- 状态轮询:根据预计耗时调整频率,避免频繁检查
- Checkpoint内容:精简关键信息,不存储完整日志
Security
- Security scanPassed, 95 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingexternal/gitcode-ascend/long-task/SKILL.md
Popularity
- Repository stars
- 128
- Repository forks
- 51
- 7-day trend score
- 24
Stars and forks are for the whole GitHub repository, not this skill alone. Refreshed 4 Jul 2026, 4 months ago.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/ascend-ai-coding/awesome-ascend-skills/long-task)