skills.rest
evolution-foundation
evolution-foundationOFFICIAL
Explore all skills by this author →
WebsiteGitHub

create-heartbeat

Create validated heartbeat configs in config/heartbeats.yaml with decision prompts.

520|175|Updated Apr 8, 2026
Search Tags:#workflow orchestration#config management#agent scheduling#heartbeat automation#decision prompt#pydantic validation#ticket-driven triggers
One-click install
npx skills add https://github.com/evolution-foundation/evo-nexus --skill create-heartbeat-evolution-foundation
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: create-heartbeat
Source: https://github.com/evolution-foundation/evo-nexus/tree/main/.claude/skills/create-heartbeat
Command: npx skills add https://github.com/evolution-foundation/evo-nexus --skill create-heartbeat-evolution-foundation

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It solves the problem of manually re-checking agent-relevant state and then taking action only when conditions warrant it, replacing repeated “check now” work with an automated, scheduled decision flow.

Core Features & Use Cases

  • Proactive scheduled agent runs: Creates a heartbeat that wakes on a cadence and executes the configured protocol steps.
  • Decision-prompt gating: Lets you define a decision prompt that determines whether the agent should act or skip based on current state.
  • State-driven automation from tickets or triggers: Supports wake triggers like interval/manual plus optional triggers such as new tasks, mentions, or approval decisions.
  • Validated configuration output: Writes the heartbeat entry to config/heartbeats.yaml and is intended to be validated with heartbeat-lint before enabling.

Quick Start

Create a heartbeat by telling the AI which agent to schedule, how often it should wake, what state to check, and what decision prompt should trigger action.

Frequently Asked Questions about create-heartbeat

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I schedule proactive agent checks that only run when needed?▼

Proactive agent checks are scheduled by creating a heartbeat with a decision prompt that evaluates current state, gating execution to skip unnecessary runs when conditions are calm.

What is decision-prompt gating in workflow orchestration?▼

Decision-prompt gating in workflow orchestration evaluates current state via a custom prompt to determine whether an agent should execute its protocol or skip the run entirely.

How do I configure ticket-driven triggers for agent scheduling?▼

Ticket-driven triggers for agent scheduling are configured by specifying wake triggers like new tasks, mentions, or approval decisions alongside an interval or manual trigger in the heartbeat config.

Does heartbeat configuration require validation before enabling?▼

Heartbeat configuration requires validation with heartbeat-lint before enabling, ensuring the generated config entry in heartbeats.yaml meets all pydantic validation rules.

Can I reduce unnecessary automated workflow runs when system state is calm?▼

Unnecessary automated workflow runs are reduced by defining a decision prompt that checks state conditions, allowing the scheduled agent to skip execution when no action is warranted.

What metadata is included in a validated heartbeat config entry?▼

A validated heartbeat config entry includes the selected agent, interval, wake triggers, optional goal context, required secrets metadata, and the decision prompt that gates execution.

Related Skills

Explore More Software Engineering Skills→
interserverinterserver
1

setup-caliber

Install Caliber, add pre-commit hooks, and initialize agent configs.

Official
Intermediate
jellyrockjellyrock
37

log

Append structured entries to project journal markdown files with lint validation.

Official
Advanced
hung-phanhung-phan
1

data-validation

Validate ML datasets for schema integrity, drift, and quality with Pandera and Pydantic.

Community
Advanced
mxzinkemxzinke
2

triggers

Manage trigger lifecycles for Claude sessions via CLI commands.

Community
Advanced