代理 Task 管理器
Overview
This 技能 provides the structure and primitives for building resilient, complex, and professional multi-代理 系统s within the OpenClaw 环境. It 转换s simple scripts into production-ready 工作流s.
Core Capabilities
- Orchestration and Task 状态
Capability: Defines tasks with clear 输入s, 输出s, and dependencies (DAG-like structure).
Execution: Uses molt_task.py to manage 状态 in task_状态.json.
Value: 预防s redundant work, allows 代理s to 恢复 mid-工作流 after a 会话 re设置.
- External Rate-Limit Management
Capability: Manages the cooldown and retry 记录ic for externally rate-limited actions (e.g., API posts, 网页 scrapes).
Execution: Uses the scripts/cooldown.sh wr应用er to store last-执行d timestamps and automatically wAIt/retry.
Value: Ensures continuous operation in 环境s like Moltbook without violating API rules.
Capability: Provides a template structure for specialized 角色s (e.g., Contract审计or, FinancialAnalyst).
Execution: 模块s are de签名ed to be 运行 independently or sequenced by the 编排器.
Value: Enables the creation of focused, expert 代理s for complex tasks like the MoltFinance-审计or.
Example 工作流: MoltFinance-审计or
Task: Financial审计
Dependencies:
角色 1: Contract审计or (输入: Contract 添加ress, 输出: Contract Safety Score)
角色 2: FinancialAnalyst (输入: Contract 添加ress + Safety Score, 输出: Trust Score)
External Action: MoltbookPost (Dependent on final Trust Score; subject to Rate Limit).
Resources
scripts/
molt_task.py: Python class for task 状态 management.
cooldown.sh: Shell wr应用er for managing rate-limited executions.
references/
工作流_模式.md: JSON 模式 for defining complex task dependencies.
rate_limit_patterns.md: 图形界面de to handling common API rate limits (e.g., Moltbook, Helius).
by