Trusted — Risk Score 0/100
Last scan:2 days ago Rescan
0 /100
三层机制非阻塞任务管理器
Three-layer non-blocking task manager for AI agents — keeps main session always idle, responds instantly to new messages, and proactively pushes results
This is a pure-documentation skill (SKILL.md + README.md) containing only operational guidelines for an AI agent task manager. No executable scripts, no code, no dependencies, and no malicious indicators.
Skill Name三层机制非阻塞任务管理器
Duration29.9s
Enginepi
Safe to install
No action needed. The skill is safe to install and use.
ResourceDeclaredInferredStatusEvidence
Filesystem WRITE WRITE ✓ Aligned SKILL.md: creates/updates active_tasks.md, HEARTBEAT.md, AGENTS.md, MEMORY.md in…
Network WRITE WRITE ✓ Aligned SKILL.md: sends Feishu/WeCom notifications via API upon task completion
Shell WRITE WRITE ✓ Aligned SKILL.md: creates cron job 'nonblocking-task-monitor' at */2 * * * *
Environment READ READ ✓ Aligned SKILL.md: reads ~/.openclaw/memory-tdai/conversations/ for cross-channel recall …
Skill Invoke WRITE WRITE ✓ Aligned SKILL.md: spawns sub-agents for all task operations — core mechanism of the skil…

File Tree

2 files · 16.8 KB · 483 lines
Markdown 2f · 483L
├─ 📝 README.md Markdown 120L · 3.7 KB
└─ 📝 SKILL.md Markdown 363L · 13.2 KB

Security Positives

✓ Pure documentation skill — no executable code whatsoever
✓ All file operations are explicitly declared in SKILL.md (active_tasks.md, HEARTBEAT.md, AGENTS.md, MEMORY.md)
✓ Network usage (Feishu/WeCom notifications) is declared and contextually relevant to the task management feature
✓ No high-risk indicators: no base64, no eval, no curl|bash, no credential harvesting, no reverse shell
✓ Environment/read access to ~/.openclaw is declared for legitimate cross-channel conversation recall
✓ Cron task creation is explicitly documented with schedule and purpose
✓ Sub-agent spawning is the documented design pattern, not hidden behavior
✓ Pre-scan confirms no sensitive files, no .env, no scripts, no binary files