Connectors, Tools, and Skills
SRE Agent: Connectors and Skills
SRE Agent retrieves log data, metrics, and runbooks from observability platforms and knowledge bases to guide responders through investigation, triage, and resolution. By connecting your observability stack and institutional knowledge, SRE Agent moves beyond simple alert analysis.
- Connectors and Tools give SRE Agent real-time access to your logs and metrics, so every triage analysis is grounded in the actual state of your systems — not just the alert payload.
- Skills give SRE Agent institutional knowledge specific to your environment — your runbooks, your escalation paths, your standards — so it can recommend precise next steps instead of generic ones.
Together, these capabilities help your team triage faster with less human involvement, reducing MTTA and MTTR while building confidence in AI-assisted operations.
Navigate to AI Settings SRE Agent Configuration to get started.
SaaS IntegrationsAll integrations described in this article are SaaS-based.
Availability
Connectors and Tools (Available in Early Access)Retrieve real-time logs and metrics by connecting SRE Agent to third-party observability platforms via API or MCP.
Skills (Available in Early Access)Arm your agent with specialized domain expertise using custom instructions, scripts, and runbooks. Skills are available through AI Settings or the PagerDuty Claude plugin.
Prerequisites
Before configuring connectors, ensure the following:
- Your account is enrolled in the SRE Agent Early Access program.
- You have Admin or Account Owner permissions in PagerDuty.
- You have credentials or API keys for the third-party integrations you plan to connect.
- The SRE Agent is enabled and configured to run via an incident workflow.
Connectors and Tools
Connectors link SRE Agent to your external data sources. Once a connector is active, the agent can use its associated tools — such as querying logs or pulling metrics — automatically or via user request. Knowledge Base integrations are invoked automatically when a URL is present in the event payload. Observability tools are user-invoked unless invoked via an incident workflow.
Supported Connectors
Navigate to AI Settings SRE Agent Configuration Connectors to view and manage your connectors.
| Connector | Tools | Connection Type | Availability |
|---|---|---|---|
| GitHub | Knowledge Base (Code — Coming Soon) | API, MCP | GA |
| ServiceNow | Knowledge Base | API | GA |
| Datadog | Logs (Metrics, Traces — Coming Soon) | API, MCP | GA |
| Grafana | Logs (Metrics — Coming Soon) | API | GA |
| New Relic | Logs | API | GA |
| Sumo Logic | Logs | API | GA |
| Splunk | Logs (Metrics, Traces — Coming Soon) | API | GA |
| AWS (CloudWatch) | Logs | API | GA |
| Elasticsearch | Logs | API | GA |
| Dynatrace | Logs | API | GA |
| Honeycomb.io | Logs, Metrics | MCP | EA |
| Observe Inc. | Logs | MCP | EA |
GA — Generally Available, EA — Early Access
Set up a Connector
- In the PagerDuty web app, navigate to AI Settings SRE Agent Configuration and click the Connectors tab.
- Find the connector you want to add and click Connect. If the connector is already listed but shows Unhealthy, click the menu to edit or reconnect it.
- On the Connect to [Integration] Tools screen, select an existing connection from the dropdown or click Manage Connections to create a new one.
- Under Agent May Use These Tools, check the tools you want to enable:
| Tool | Description |
|---|---|
| Logs (API) | Allows SRE Agent to query log data from this source during incident triage. |
| Metrics (MCP) | Allows SRE Agent to pull metric data. Available on connectors that support it. |
- Click Save. The connector appears in your list with a status of
Activeonce the connection is established.
Connector Statuses
- Active: The connector is healthy and SRE Agent can use its tools.
- Unhealthy: The connection exists but is failing. Check your credentials or API key and reconnect.
- Not Connected: No connection has been set up. Click Connect to configure it.
Skills
Skills allow you to arm SRE Agent with custom instructions, runbooks, and scripts that give it specialized domain expertise. Rather than relying solely on retrieved data, a skill tells the agent exactly how to handle specific situations, so it can intelligently deduce the next steps for troubleshooting issues even before a human looks at the incident.
What a Skill Contains
Each skill is a structured set of instructions with the following components:
| Component | Details |
|---|---|
| When to Use | Conditions that tell the agent when to invoke this skill automatically. |
| Prerequisites | The integrations or data the agent needs to execute the skill. |
| Execution Steps | Step-by-step instructions the agent follows during triage. |
| Error Handling | How the agent should gracefully degrade if a step fails. |
| Success Criteria | How the agent determines whether the skill ran successfully. |
| Examples | Trigger phrases or alert conditions that invoke this skill. |
Example Skill: detect-security-incident-logs
The following is an example of a fully structured skill that you can use as a template. This skill automatically searches Datadog and AWS CloudWatch logs for critical security events during incident triage.
Name: detect-security-incident-logs
Description: Searches Datadog and CloudWatch logs for critical security incidents. When SECURITY_INCIDENT: CRITICAL log entries are found, it recommends triggering the Alert Security Team incident workflow.
Example Instructions:
When to use:
- Automatically during initial triage when the alert custom details include type: possible_security_issue
- When a responder asks the agent to check for security incidents
Prerequisites:
- Datadog connector configured and active
- AWS CloudWatch connector configured and active
Execution steps:
1. Search Datadog logs for the exact phrase SECURITY_INCIDENT: CRITICAL within the incident time window.
2. Search AWS CloudWatch logs for the exact phrase SECURITY_INCIDENT: CRITICAL within the incident time window.
3. If matches are found in either source, surface the key log details in the SRE Agent response.
4. Recommend triggering the Alert Security Team incident workflow.
Error handling:
- If one connector is unavailable, continue with the other and note which source was skipped.
- If both connectors are unavailable, inform the responder and suggest checking logs manually.
Success criteria:
- The agent has queried both log sources and reported results or confirmed no matches were found.
Examples:
- Alert custom_details contains: type: possible_security_issue
- Responder asks: "Check for any security incidents in the logs"
Create a Skill
Skill can be created through the PagerDuty UI or with the PagerDuty Claude plugin.
In the PagerDuty Web App
- In the PagerDuty web app, navigate to AI Settings SRE Agent Configure and click the Skills tab.
- Click Add Skill.
- On the New Skill page, complete the following fields:
| Field | Value |
|---|---|
| Name | Enter a unique identifier for the skill (for example, escalation-runbook). |
| Description | Enter a short description of what the skill does (for example, Guides the agent when escalating incidents). |
| Instructions | Enter the full instructions to send to the LLM. Use the structure described in What a Skill Contains as a guide. |
| Make skill available to everyone | Check this option to make the skill accessible to all users in the account. When unchecked, the skill is only available to you. |
- Click Save. The skill is immediately available to SRE Agent for all future incidents.

Create a skill in PagerDuty
Using the Claude Plugin
- Install the PagerDuty Claude plugin.
- Run the
/pagerduty:create-pagerduty-skillcommand and follow the guided prompts to define the skill's name, instructions, trigger conditions, and metadata. - Once created, the skill is immediately available to SRE Agent for all future incidents.
Test a Skill
After creating a skill, trigger an incident that matches your skill's conditions and review what the agent reports in:
- The SRE Agent tab on the incident details page.
- Your ChatOps channel (Slack or Microsoft Teams).
Updated 3 days ago
