Changelog

​

Incident action items and follow-ups

• Two types: action items (Action) attach to the incident and capture to-dos during response; follow-ups (Follow-up) attach to the post-mortem and track remediation after the review
• Each work item has a title, description, status, priority, and assignees, and the list can be filtered by type and assignee
• Assignees-only completion: assignees can mark a work item complete
• After the post-mortem, an action item can be converted into a follow-up, binding automatically to the incident’s post-mortem
• Creation, updates, completion, conversion, and deletion are all recorded on the timeline
• Available to On-call Pro and above accounts

​

Comment types

• Account-level configuration: create, rename, recolor, reorder, and delete, up to 10 types per account
• Type names up to 40 characters, with custom colors (#RRGGBB)
• Pick a type when publishing a comment; the timeline shows the chosen type

​

Entity Tree troubleshooting views

• Runtime status: see the current availability state (Healthy / Initializing / Unavailable / Unsupported version / Unknown) and its reason, the responsible Edge instance (ID / IP / version / last heartbeat), and the center-vs-Edge plan digest comparison
• Deployment plan (debug): inspect the execution plan compiled from the saved configuration — data source bindings, groups and matcher conditions, and per-rule, per-severity policies with PromQL; template, plan, and rule digests are one-click copyable for support tickets. The full plan is visible only to entity-category managers
• Query preview: run the condition or recovery PromQL against a selected data source right in the rule editor and read diagnostics such as returned rows, valid entities, and missing-identity, duplicate-alert, and scalar-conflict warnings
• Execution history: review the last 24 hours of actual evaluations for each rule — the executed PromQL, outcomes, and error summaries; requires VictoriaLogs and caps the display at 200 records

​

Alert Rule Experience Upgrade

• Per-rule runtime details: the alert-rule list adds a Runtime Details entry showing rule runtime status (Normal / Abnormal / Awaiting first execution / Stale), the last execution time and duration, active-alert instance breakdown (Threshold / AnyData / NoData), and the last 24 hours of execution history (requires VictoriaLogs, capped at 200 records)
• Import/export redesign: import is now a file upload that auto-detects four formats — Flashduty Rules YAML, legacy Flashduty JSON, Prometheus Rules YAML, and Nightingale v6–v9 rule JSON (single file, up to 1MB) — with a conflict policy (skip same-name rules / create copies), pre-import preview, and per-rule selection; export produces a YAML bundle (default filename alert-rules.yaml) for cross-tenant migration
• Bulk metadata editing: extra labels and custom fields support incremental patch (per-key set/delete with value suggestions and impact preview; suggestions disabled beyond 500 selected rules) and full replacement modes, with failures reported per rule
• Active-alert details: active alerts gain a per-row detail drawer (overview, timing, rule and source, current values, alert dimensions, event labels, alert annotations, diagnostics) and an Alert Rule filter; clicking a rule’s status cell jumps straight to that rule’s active-alert view
• Recovery and end modes: Threshold and AnyData detection add a Manual close recovery mode (the alert stays active until closed by hand); no-data alerts gain an End mode (auto-end when data reappears / auto-end on recovery or timeout with a configurable auto-close time / manual close only)
• Custom-field templating: custom field values support ${label} variables and Go templates (a value containing {{ is rendered like the description template), and a field named title becomes the event title

​

Status Page Widgets

• Two forms: the status badge always shows the current overall status — ideal for footers and help centers; the event banner appears at the top or bottom of the page only during an incident or maintenance, can be dismissed by visitors, and shows scheduled maintenance automatically 24 hours before it starts
• Configurable appearance: theme (Auto / Light / Dark), language (中文 / English), badge size, and banner position
• Easy to embed: go to your status page detail view → Settings → Widget, preview live, and copy the embed code in one click (<script> + the <flashduty-status-widget> Web Component)
• Public JSON API: GET {status page URL}/api/widget/v1/summary.json — no key required, with CORS and ETag support, so you can render the data yourself or integrate it into your monitoring stack
• Available on public status pages only; enabled by default with no configuration needed

​

Artifacts can now be shared with a public link

• The share panel offers two visibility modes, Account only and Public link; the default is still account-only
• A public link shows a content snapshot taken when the link was generated and does not sync automatically. When the artifact changes, click Update snapshot to sync it — the link itself stays the same
• Revoke public link at any time; the link stops working immediately
• Artifacts larger than 16 MiB cannot generate a public link

​

ServiceMap (Beta)

• Filter the topology canvas by resolution status (Confirmed / Candidate / Unresolved), adjust the hop range, and double-click a node to enter focus mode
• Node and dependency detail panels expose identity, runtime, resolution, and observation fields, so you can judge whether a dependency is trustworthy
• Unresolved endpoints are grouped by reason, with filtering and CSV export
• The ServiceMap host list filters by Agent version, Edge cluster, capture mode, and status
• Query info reports capture mode, evidence freshness, and truncation reasons, so you can tell whether the current topology is complete

​

AI SRE Open to All Accounts

• Removed the beta application form and gradual-rollout whitelist; the AI SRE entry is visible to all accounts by default
• Entry and actions follow the role permissions configured in your account
• Flashduty will share billing information ahead of general availability

​

AI SRE autonomous investigation Agent

• Start investigation sessions from the console chat workspace and inspect streaming output, tool calls, and conclusions
• Bring incident or war room context into AI SRE sessions so the Agent investigates a specific incident
• Extend investigation capabilities with Skills, Knowledge Packs, MCP, and A2A Agents
• Use /insight to review the last 30 days of AI SRE sessions, repeated context, and missing runbooks

​

IM-native investigation and automatic war room diagnosis

• Reply in IM threads to keep investigation discussions focused
• Automatically run an initial diagnosis when a war room is created and post the result back to the room
• Use /env to switch the Environment bound to the current IM session
• Use /scope to switch the team scope bound to the current IM session

​

Automation and BYOC Runner

• Trigger Automations with cron schedules, HTTP POST, or On-call incidents
• Use preset templates, run history, manual runs, and read-only permission controls
• Choose automatic Environment selection, cloud Sandbox, or a self-hosted BYOC Runner
• Run the Runner with Linux systemd, Docker, or manual mode, and constrain command execution through a local permission config

​

WeChat Mini Program RUM

• Automatically collect page lifecycle events, user actions, network requests, application errors, and performance metrics
• Configure service, env, version, session sample rate, and proxy reporting
• Use the new WeChat Mini Program insights dashboard for UV, sessions, errors, launch time, first render, and setData metrics
• Analyze performance trends by version, environment, loading type, and operating system

​

HarmonyOS SDK

• Use the @flashcatcloud/core, @flashcatcloud/rum, @flashcatcloud/trace, and @flashcatcloud/crash modules
• Collect views, user actions, network requests, errors, and crash events
• Inject Trace context through the rcp interceptor or the FlashcatHttp wrapper
• Follow guidance for HarmonyOS SourceMap and native symbol file uploads

​

Mobile symbolication and compliance guide

• Upload WeChat Mini Program sourcemap.zip files and restore mini program stacks
• Upload Android ProGuard/R8 mapping files and NDK native symbol files
• Upload iOS dSYM symbol files
• Use the new SDK developer compliance guide for privacy policy disclosure, delayed initialization, collected fields, and data that the SDK does not collect
• Use the new Web SDK performance impact guide for SDK size, CPU, memory, network overhead, and Session Replay sampling recommendations

​

HTTP Pull and DB Pull alert integrations

• HTTP Pull: Periodically call an external HTTP endpoint, with support for GET/POST, headers, request bodies, timeouts, retries, cursor pagination, and severity mapping
• DB Pull: Periodically query MySQL, PostgreSQL, or ClickHouse and convert rows into standard alert events through field mappings
• DB Pull uses Keyset Pagination with time and ID cursors to pull incrementally and avoid full table scans
• Both integrations support dedicated and shared integration modes, and can route events into target channels through default routing and routing rules

​

Dynamic assignment append mode

• Use reset mode to replace targets in the template escalation rule
• Use append mode to add notification targets, useful when you want to keep a default team and add business owners
• Combine label mapping with CMDB, configuration platforms, or CSV data to generate dynamic assignment labels automatically

​

CLI, Go SDK, and Open API

• Flashduty CLI manages incidents, changes, members, teams, channels, status pages, and notification templates from the terminal
• CLI supports table, json, and toon output formats, plus install, update, and mirror download configuration
• The Go SDK wraps the Open API in a go-github-style typed client, covering 288 API operations across 32 services
• The Open API reference continues to add AI SRE, RUM, On-call, and Platform endpoints

​

RSS/Atom Feeds for Public Status Pages

• Every public page serves four URLs: feed.rss, feed.atom, history.rss, and history.atom. The history.* pair is an alias for Atlassian Statuspage compatibility
• Feeds include incidents and maintenance events published in the last 90 days, up to 50 items, ordered by start time descending
• Hidden sections and components are automatically filtered out and do not appear in the feed
• Responses support ETag / If-None-Match and return 304 Not Modified on cache hits

​

Team Filter on the Incident List

​

New Alert Events Counter on Incident Rows

​

AI Post-Mortem Report

• One-click Generation: AI automatically distills incident overview, impact scope, root cause analysis, and improvement recommendations
• Collaborative Editing: Multiple team members can edit the post-mortem report online with real-time sync
• Access Control: Set access permissions for post-mortem reports to protect sensitive information
• Image Support: Upload and embed images directly in post-mortem reports

​

AI Assistant

• Instant Q&A: Chat with AI directly in the console to quickly get incident investigation suggestions
• Mermaid Diagram Rendering: AI-generated flowcharts and architecture diagrams are rendered visually with zoom and fullscreen support
• Image Paste: Paste screenshots directly from your clipboard for easy context sharing
• Document Search: Intelligent retrieval and answers based on the Flashduty knowledge base

​

External Incident Submission

• Submit incidents via API or a standalone page
• Upload screenshots and attachments with submissions
• Automatically match escalation rules for incident assignment

​

SSO Subdomain Login

• Account administrators can configure a dedicated login subdomain
• Supports automatic account switching based on subdomain
• SSO-synced users can be set as non-editable to prevent manual changes from overriding SSO data

​

Other Optimizations

1. Manual Incident Metadata Override: Manually modify incident title, description, and severity
2. Voice Notification Templates: Voice notifications now support customizable templates instead of fixed formats
3. Webhook Custom Payload: Incident webhooks support user-defined request body formats
4. Push Comments by alert_key: New API to push comments to incidents via alert key
5. War Room Message Limit Increase: Increased war room message capacity for more complete incident communication records

​

New Alert Integrations

• JD Cloud: Support for JD Cloud monitoring alert ingestion
• KS Cloud: Support for KS Cloud monitoring alert ingestion
• UCloud: Support for UCloud monitoring alert ingestion
• Harbor: Support for Harbor container registry alert ingestion
• Nagios: Support for Nagios monitoring alert ingestion

​

Monitors Active Alerts

• Browse active alerts by folder
• Flexible filtering based on labels
• Prometheus-compatible label values query API (/api/v1/label/:label/values)

​

Other Optimizations

1. Custom Action Fields Expanded: Custom webhook actions support additional fields
2. Feishu War Room Chat History: Retrieve chat message history from Feishu war rooms
3. Webhook Secret Verification Skip: Webhook integrations can optionally skip secret verification
4. Custom Webhook Retry: Custom webhooks support configurable retry conditions and automatic retry

​

RUM Mobile Source Map

• After uploading mobile source map files, error stacks are automatically resolved to readable source code locations
• Supports iOS dSYM and Android Proguard/R8 mapping files

​

RUM Error Ingestion Exclusion Rules

• Set exclusion rules by error type, message content, and other conditions
• Excluded errors no longer count toward ingestion volume

​

Issue Preset Severity

​

Monitors Edge Enhancements

• VictoriaLogs Support: New VictoriaLogs data source with raw query capability
• Fluent-bit Integration: New Fluent-bit data source support
• Loki Query Optimization: Support loki.start/loki.end time parameters as first priority, raw query mode
• Template Function Expansion: Alert rule templates gain timeFormat function and the Sprig template function library

​

Flashduty MCP Server V2

• Streamlined tool set for improved interaction efficiency
• New /mcp endpoint
• Structured logging with distributed tracing support

• labels.application_name = "order-service"
• Alert will be automatically routed to the channel named order-service

• Support drag-and-drop sorting of routing rules for flexible priority adjustment
• Optimized rule content display with clearer conditions and actions
• Cleaner overall layout, reducing configuration and troubleshooting costs

​

War Room

• Instant Group Creation: One-click creation of Feishu/Lark, Dingtalk, WeCom, or Slack group chats, automatically inviting responders and stakeholders
• Message Sync: Incident message cards are automatically delivered to the group, always pinned and updated with the latest status in real-time

​

Webhook Call History

• Call status and error codes
• Retry count
• Complete request and response information

​

Change History

• Dynamic Routing: Change events can be automatically routed to relevant channels based on business
• Label Enhancement: Support secondary processing of reported events
• New View: Change history page upgraded with custom view support

​

Other Optimizations

1. Incident List: Support batch assignment of incidents
2. Outlier Incidents: Support disabling outlier incident detection
3. Alert Storm Reminder: Support configuring multiple storm thresholds for tiered progressive reminders
4. Escalation Rule: Group chat channel configuration can specify application severity
5. Incident Details: Support custom label sorting, support displaying incident labels in JSON format
6. Integration Center: Added Zoho ServiceDesk Plus and Cloudflare integrations

​

AI Summary

1. Event overview (key information)
2. Impact scope (resources, services, etc.)
3. Actionable measures (investigation, mitigation, prevention)

​

ServiceNow Integration

1. Support manual and automatic sync triggers
2. Support bidirectional sync between Flashduty and ServiceNow

​

Mobile App Optimization

1. Added English version support
2. Integrated AI Summary feature
3. Incident details now include custom fields, optimized label display and text editing experience

​

Insights

​

Microsoft Teams

​

WeCom/Zoom/Feishu Bot

1. Support uploading email and account ID mappings for Feishu/Lark and other platforms
2. Escalation rules support binding mappings for precise @ mentions in group chat pushes

​

Link Integration

​

Dingtalk and Feishu/Lark Alert Integration

​

Incident Merge Interaction Optimization

​

Intelligent Alert Grouping Supports Custom Fields

​

Incident Auto-close Timeout Optimization

​

Shared Integration Permission Management

​

Incident Convergence Upgraded to Flapping Detection

​

License Management

1. New Fixed License Type: Permanently valid, cannot be preempted
2. Flexible License Assignment: Account administrators can set fixed or temporary licenses for members
3. Member Self-management: Regular members can voluntarily release their own license
4. API Batch Operations: Support batch setting member license types via API

​

Routing Version Management

1. Historical Version Viewing: Integration center alert routing now supports viewing historical versions
2. Version Comparison: Support comparing current routing rules with historical versions
3. One-click Rollback: Support rollback to any historical version of routing rules

​

Jira Integration

• Manually or automatically sync incident information to Jira Issues
• Support both Jira Cloud and Server versions
• Sync incidents within a limited scope
• Incident field mapping management

​

Template Management

• New Interaction Design: Template management interface redesigned
• Historical Incident Preview Debugging: Support selecting historical incidents for template preview and debugging
• Smart Input Hints: System automatically provides suggestions after typing {{

​

Other Optimizations

• IM Integration: Added support for Feishu/Lark and Dingtalk private deployment versions
• Incident Webhook: Support subscribing to incident comment events
• Incident Convergence: New incident convergence feature, can be enabled or disabled

​

Incident Management at Your Fingertips

• Full Mobile Workflow: Support incident search, detail viewing, acknowledgment, closing, escalation, and manual triggers, improving key operation response speed by 40%
• Multi-dimensional Information Aggregation: Incident timeline, processing records, and associated logs displayed together

​

Important Notifications, Never Missed

• iOS Critical Alerts: Based on Apple’s official Critical Alerts protocol, breaking through Do Not Disturb mode restrictions
• Android System-level Channels: Support system-level channels for all major Android manufacturers

​

Voice Notifications, More Reliable

• Smart Whitelist Sync: Automatically add Flashduty voice numbers to phone contacts

​

Added Metric Viewing Dimensions

​

Time Period Segmentation

• Work hours: Weekdays 8am-7pm
• Rest hours: Weekdays 7pm-11pm, Non-weekdays 8am-11pm
• Sleep hours: Every day 11pm-8am

​

New Workload Metrics

• Interruptions: High-priority messages like SMS, phone calls, and app pushes are marked as interruptions
• Response Effort: Sum of time between member acknowledging incident and incident resolution

​

Report Download and Data Export

1. New Alert Processing Feature: Support multiple execution actions
2. Optimized Label Enhancement and Routing Rules Configuration: Added right-side alert history panel to assist rule debugging

​

Channel Access Level

​

My Channels Filter

​

Data Isolation Best Practices

• Personnel Isolation: Register multiple owners directly, with complete isolation of personnel and resources
• Resource Isolation: Set channel access level to private
• Visual Experience: Keep incident list always filtered to my channels

​

On-call Rotation Notifications Support Scheduled Notifications

​

On-call Rotation Notification Content and Style Optimization

​

New Date Mask Mode

​

Other Optimizations

1. Schedule preview adds 2-week mode
2. Schedule preview calendar mode, color changed from following rules to following rotation
3. Added Highlight My On-call button

1. Browser Language Setting: All members can now freely switch browser display language in the console
2. Notification Language Setting: Account owners can select the default notification language in “Account Settings”

​

CTYun Alert Integration Support

​

Guance Alert Integration Support

1. Brand new UI interaction design
2. New incident list page with rich filtering support
3. New incident details page supporting common actions and custom actions
4. New incident creation support with escalation rule and direct assignment
5. New dark theme switching support

​

Intelligent Grouping

​

Outlier Incidents

​

Past Incidents

• Matched silence or inhibit rules before incident assignment notification, blocking incident notification if matched

• Match silence or inhibit rules during alert event delivery, blocking incident generation if matched
• Blocked alerts will no longer trigger or merge into incidents
• You can choose to retain or discard blocked alerts

1. Console requires secondary confirmation before clicking custom actions
2. Newly created channels displayed at the top
3. Create incident page adds escalation rule specification, supports markdown description
4. Assignment records display dynamic assignment reset indicator
5. Video demo entries added near core feature points
6. Escalation rule single chat unified configuration supports not setting channel for certain severity levels
7. Incident reassignment removes current responder

1. Support Zabbix v7 version alert integration
2. Support InfluxData v2 version alert integration
3. Support Meraki alert integration
4. Support Zoho OpManager alert integration