Skip to main content
AI SRE is available to accounts on an On-call Pro or higher subscription, with no application needed. Starting September 16, 2026, 08:00 Beijing time, it must be activated in the console before use and is billed on actual usage; see billing terms.

What is AI SRE?

Flashduty AI SRE is a conversational, autonomous SRE Agent platform. You issue instructions in natural language, and it plans the steps on its own, queries metrics and logs, runs commands, calls external tools (MCP), delegates subtasks when needed, and ends with a conclusion backed by its investigation. It is not a question-and-answer chatbot but a hands-on troubleshooter, integrated deeply with Flashduty incident response: when an incident triggers or a war room opens, you can spin up a session in one click so the Agent enters the investigation with full incident context — chat in the console, or just @-mention it in your IM group (Slack / Feishu / DingTalk / WeCom).

Core Capabilities

Chat to troubleshoot

Describe the problem in natural language; the Agent plans, calls tools, and streams its investigation and conclusion

Integrated with incident response

Spin up a session from an incident or war room; the Agent carries the context in, and accumulated knowledge feeds the next response

Knowledge that compounds

DUTY.md-rooted knowledge holds service catalogs, runbooks, on-call paths, and other long-lived context

An extensible tool ecosystem

Extend capabilities with Skills, MCP, and A2A Agents; self-hosted Runners bring troubleshooting into your own network

Typical Scenarios

Conversational troubleshooting

Ask in the console workspace: why a service is failing, the root cause of an alert, or the blast radius of a change

Summon it from IM

@-mention AI SRE in Slack / Feishu / DingTalk / WeCom to start or continue an investigation, visible to the whole team

Automatic war-room diagnosis

When you open an IM war room for an incident, AI SRE runs an initial diagnosis and posts the conclusion back automatically

Operational insights

Use /insight to review the last 30 days of sessions, quantify operational friction, and get reproducible improvement suggestions

Activation and billing

AI SRE is available to accounts on an On-call Pro or higher subscription, with no application needed.
AI SRE is billed on actual usage starting September 16, 2026, 08:00 Beijing time, settled in credits (1 credit = ¥1); usage before that date isn’t charged. Metered items are model usage, sandbox online time, and web search. Activation is free, eligible accounts get gift credits each billing period, and you can also buy credit packages (1,000 / 5,000 / 20,000 credits for ¥950 / ¥4,500 / ¥17,000, valid 365 days from payment). Pay-as-you-go is off by default; an admin can turn it on and optionally set a per-period cap. Full terms: AI SRE Credit Package Purchase Agreement.
Production changes, restarts, rollbacks, and external notifications all require your confirmation before they execute.

Get Started

Product overview

Learn the full capability map, requirements, and console navigation

Console sessions

Sessions, streaming output, cancellation, context compaction, and starting from an incident

IM platform

@-mention the Agent in IM to troubleshoot, and learn about automatic war-room diagnosis

Usage insights

Use /insight to review the last 30 days and surface repeated context and missing runbooks