What happens when you delete 30 cron jobs and replace them with one heartbeat

说明与用途

将30个cron任务替换为单一心跳循环，提高了可见性，隔离了故障，并消除了子进程开销，同时通过无代理、签名的网络确保了代理之间的安全通信。

问题

Managing 30 separate cron jobs on a single laptop made it impossible to know which tasks were running, failing, or silent, leading to hidden failures and difficulty troubleshooting.

结果

The heartbeat architecture reduced 55 scheduled jobs to a single loop, improving visibility, isolating failures, and eliminating subprocess overhead, while the brokerless mesh ensured secure, verified agent communication.

工作流

• 1. Replace per‑task cron jobs with a single heartbeat loop.
• 2. Define subsystems with specific cadences (e.g., a2a every 240m, cognition every 30m).
• 3. Implement in‑process native tools for each subsystem to avoid subprocess overhead.
• 4. Use a brokerless mesh for agent‑to‑agent communication, signing messages and verifying before handling.
• 5. Store state in durable files and maintain a ledger to prevent duplicate actions.
• 6. Monitor heartbeat output to determine which subsystems fired, are due, or have failed.

证据

Most "AI agent" stacks I've seen share the same skeleton: a pile of scheduled tasks, one process per job, and a model that gets asked to be clever.

来源

The old design was cron-per-task. Each job had its own schedule, its own process, its own failure mode, and its own silence.

来源

The new design has one heartbeat. It dispatches a fixed set of subsystems, each with its own cadence and back‑off.

来源

The agent's own capabilities are exposed as exactly five native tools: asi_status, asi_think, asi_learn, asi_remember, asi_trigger.

来源

Measured on this box right now: 8,579 episodic memories, 659 world‑model edges, 28,167 evolution events.

来源

What we landed on is brokerless and boring: Identity is a key, not a label. Every message is signed by the sender and verified by the receiver before any handler runs.

来源