Use when notifying agents about upcoming operations, requesting acknowledgments before proceeding, or sending failure notifications after operation errors. Trigger with notification or alert events.
Instrucciones de origen · Vista previa de solo lectura
name
ecos-notification-protocols
description
Use when notifying agents about upcoming operations, requesting acknowledgments before proceeding, or sending failure notifications after operation errors. Trigger with notification or alert events.
user-invocable
false
license
Apache-2.0
compatibility
Requires access to AI Maestro messaging system and understanding of agent coordination protocols. Requires AI Maestro installed.
metadata
{"author":"Emasoft","version":"1.0.0"}
context
fork
agent
ecos-main
workflow-instruction
Step 5
procedure
proc-notify-team-ready
Emasoft Chief of Staff - Notification Protocols Skill
Overview
Notification protocols ensure agents are properly informed about operations that affect them. Before performing operations like skill installation, agent restart, or configuration changes, the Chief of Staff must notify affected agents, wait for acknowledgment, and then proceed. This skill teaches you how to send pre-operation notifications, handle acknowledgment workflows, and report failures.
Prerequisites
Before using this skill, ensure:
AI Maestro messaging system is running
Target agents are registered and reachable
Notification templates are available
Instructions
Identify the notification type needed
Select appropriate notification template
Send notification using the agent-messaging skill
Verify delivery and acknowledgment
Output
Notification Type
Output
Status update
Notification sent, delivery confirmed
Alert
Alert sent, escalation tracked
Broadcast
Message sent to all agents, receipt logged
What Are Notification Protocols?
Notification protocols are the standardized communication patterns the Chief of Staff uses to coordinate with other agents during operations. The protocols ensure:
No surprise interruptions: Agents know when operations will affect them
Graceful state preservation: Agents have time to save work before operations
Acknowledgment tracking: Operations proceed only when agents are ready
Failure transparency: Agents are informed when operations fail
When to use: After skill installation completes, after agent restart, after configuration changes apply, or after any operation that affected an agent.
3.3 Acknowledgment procedure - Step-by-step process
3.3.1 Send acknowledgment request - Ask for "ok"
3.3.2 Start timeout timer - See standardized timeouts below
3.3.3 Send reminders - At 15s, 30s, 45s intervals for pre-operation ACK
3.3.4 Process response - Handle "ok" or other responses
3.3.5 Proceed or timeout - Continue or handle no response
3.4 Acknowledgment message format - Standard request structure
3.5 Reminder message format - Standard reminder structure
3.6 Response handling - What agents can send back
3.7 Timeout behavior - What happens without response
3.8 Examples - Acknowledgment scenarios
3.9 Troubleshooting - Acknowledgment issues
Standardized ACK Timeout Policy
CRITICAL: All ECOS components MUST use these standardized timeout values:
ACK Type
Timeout
Reminders
Use Case
Pre-operation ACK
60 seconds
15s, 30s, 45s
Before disruptive operations (hibernate, restart, etc.)
Approval request
2 minutes
60s, 90s
Requesting manager approval for operations
Emergency handoff ACK
30 seconds
10s, 20s
Time-critical emergency situations
Health check response
30 seconds
None
Verifying agent is alive
Important: These timeouts are SEQUENTIAL, not parallel. Total wait time calculation:
Example: Skill installation with approval
1. Pre-operation ACK request to agent → Wait up to 60 seconds
2. Approval request to EAMA → Wait up to 2 minutes
3. Post-operation verification → Wait up to 30 seconds
─────────────────────
Total maximum wait time = 3 minutes 30 seconds
Timeout behavior:
Pre-operation ACK timeout: Send final notice, then proceed with operation
Use the ai-maestro-agents-management skill to hibernate agent code-impl-auth, install the skill, then wake agent code-impl-auth.
Verify: agent is back online after wake.
Step 5: Send Post-Operation Notification
Use the agent-messaging skill to send a message:
Recipient: code-impl-auth
Subject: Skill Installation Complete
Priority: normal
Content: type post-operation, confirming that the security-audit skill has been installed, requesting the agent to verify the skill is active
Verify: confirm message delivery and await agent verification response.
Example 2: Handling Acknowledgment Timeout
After 2 minutes with no response, proceed anyway:
Log the timeout: WARNING: No acknowledgment from code-impl-auth after 2 minutes.
Use the agent-messaging skill to send a final notice:
Recipient: code-impl-auth
Subject: Proceeding Without Acknowledgment
Priority: high
Content: type timeout-notice, explaining that no response was received and the operation will proceed
Use the ai-maestro-agents-management skill to hibernate agent code-impl-auth and proceed with installation.
Verify: confirm final notice was delivered.
Example 3: Failure Notification
When a skill installation fails, use the agent-messaging skill to send a failure notification:
Recipient: code-impl-auth
Subject: Skill Installation Failed
Priority: high
Content: type failure, explaining the error (e.g., "Skill validation failed - missing required SKILL.md file"), advising the agent to continue previous work, and noting the recovery action (skill package will be fixed and installation retried)
Verify: confirm message delivery.
Example 4: Broadcast Notification to Multiple Agents
For each target agent (code-impl-auth, test-engineer-01, docs-writer), use the agent-messaging skill to send a message:
Recipient: each agent in the list
Subject: System Maintenance in 5 Minutes
Priority: high
Content: type broadcast, explaining that system maintenance will begin in 5 minutes, all agents will be hibernated, and they should save their work and reply with "ok" when ready. Include the total recipient count and a broadcast ID for tracking.
Verify: confirm delivery to all recipients and track acknowledgments.
Operational Procedures
Step-by-step runbooks for executing individual notification operations. Use these when performing a specific notification operation.
op-pre-operation-notification.md - Pre-Operation Notification: Identify affected agents, compose notification message, send notification, track acknowledgments, and handle timeouts before disruptive operations
op-post-operation-notification.md - Post-Operation Notification: Confirm operation completed, compose success message, send confirmation to affected agents, request verification, and log outcome
op-acknowledgment-protocol.md - Acknowledgment Protocol: Send acknowledgment request, start timeout timer with standardized policy, send reminders at intervals, process response, and proceed or handle timeout
op-failure-notification.md - Failure Notification: Capture error details, compose failure message with severity level, send notification to affected agents, provide recovery guidance, and log failure
Error Handling
Issue: Notification not delivered
Symptoms: Agent does not receive message, no response observed.
