New App
- Added the CAST AI audit log collection app
- Periodically calls the CAST AI audit events API (`GET /v2/audit/events`) to collect audit logs and normalizes them with the `cast-ai-audit` log schema
- Logpresso Sonar 4.0.2502.0 or later is required
Logger
- Added the `CAST AI Audit Logs` logger model
- Scan days: number of days to look back on the first collection (1 to 90, default 1)
- Collection delay (seconds): time to postpone collection to allow for the delay before events become available in the API (0 to 3600, default 2)
Connect Profile
- Added the `CAST AI` connect profile type
- Configure the API key, API URL (default / EU / India regions), HTTP proxy, and connect and read timeouts
- Connection test supported
Query Commands
- Added `cast-ai-audit-events`: retrieves audit events directly from the CAST AI server
- Options: `profile`, `from`, `to`, `limit`
- Returns the same field names as the logs stored by the logger (such as `event_id` and `actor_email`)
- Can be used with `import` to backfill events the logger missed due to the collection delay
Log Schema
- Added the `cast-ai-audit` log schema
New Features
- Added `WebKeeper SG 넷앱 (Elastic)` logger model for Elastic NetApp logs
- Added `webkeeper-netapps-logs` query command for NetApp logs
- Added `WebKeeper SG (Syslog)` logger model for web and NetApp logs
Improvements
- WebKeeper query commands are listed in playbook commands
Changes
- Renamed Elastic web logger model: `WebKeeper SG` → `WebKeeper SG 웹 (Elastic)`
- `_time` of file (SFTP) logs is now the event time instead of the collection time
- Added `agent_id`, `category_code`, `service_code`, `user_guid` fields to the `webkeeper-netapps` log schema (Elastic logger only)
Bug Fixes
- Fixed web log collection and query failing when NetApp logs are mixed in the Elastic index
- Fixed Elastic logger bugs (connect profile changes were applied only after saving the logger again, possible log loss when collection was interrupted)
- Fixed `webkeeper-logs` query command bugs (`offset` and `limit` applied again for each index, failure in distributed queries)
Upgrade Notes
- File (SFTP) logs collected with 1.2.2610.0 have the collection time in `_time`, so their time basis differs from data collected with 1.3.2611.0 or later
Bug Fixes
- Fixed an issue where installing or upgrading the app failed with `duplicated connection profile factory type: slack-webhook`
Notes
- Using the app as the AI agent guardrail requires the AI Assistant or Agentic SOC app 1.13 or later.
New Features
- AIM Starfort connect profile
- Guardrail Test screen
- AI agent guardrail provider
- Query commands: starfort-guard, starfort-guard-batch
Improvements
- Support for the SIEM log format of Endpoint Protector 2608 or later
- Changed log field names are normalized to the existing fields, so existing log schemas and dashboards work as before
- `Policies Received` and `User Login` logs are classified as client audit logs
- Policy and detected item fields in content protection logs (multiple detected items are joined with commas)
Bug Fixes
- Fixed the `event_time` year of Syslog logs received after the year changes being stored as the year the app started
Improvements
- `event_time` of December Syslog logs received in January is set to the previous year
Changes
- Syslog logs with a nonexistent date (e.g. Feb 30) are no longer corrected to another date and fail to parse
Improvements
* Minimum Memo Count Filter: Added a condition to sync only tickets with at least *N* memos (requires both instances to be on 2.6.2611.1+; otherwise stops with a `REMOTE_OUTDATED` error).
* Per-Mapping Notification Groups: Automatically assigns a mapping's configured notification group to newly created placeholder detection scenarios.
* Batch Limits & Intermediate Commits: Capped per-cycle processing at 200 items by default and added checkpoint commits every 100 items to prevent thread blocking and lost progress.
* Immediate Memo Sync with Tickets: Fetches memos immediately after a ticket is synced, resolving delays and missing pre-watermark memos.
* Other Improvements: Added raw event counts and one-line summaries to evidence tracking logs, fixed a numeric input clearing bug in settings, and moved long descriptions into `?` tooltips.
※Upgrade Note: To use the Minimum Memo Count Filter, both the collector and remote instances must be upgraded to 2.6.2611.1 or higher. All other features work by upgrading the collector side only.**
New Features
- Added t he `Google Gmail Activity` logger
- Added the `google-workspace-gmail-logs` query command
- Added the `Google Gmail Activity` log schema
- Added a Gmail tab to the dashboard
- Uses the same permissions as existing audit logger
Notes
- Gmail logs have limitations due to what Google provides. See the install guide
New Features
- Query command that retrieves issue activities (`jira-issue-activity`)
Improvements
- Page Size and Slice Minutes options for the audit log and issue activity loggers and the `jira-audit-logs` query command
- Profile option for the `jira-audit-logs` query command
- Optimized issue activity logger query logic (fewer API calls)
- Automatic retry on temporary Jira server errors (HTTP 502, 503, 504)
Changes
- Default read timeout of the `Jira` connect profile changed from 30 to 60 seconds
Changes
- Added the Log start time option to the Reporting API loggers (applies to newly created loggers)
- When more logs than the API can page through arrive within one second, the Reporting API logger skips the excess and writes a warning log
Bug Fixes
- Fixed the Reporting API logger storing logs twice in the second on a slice boundary
Upgrade Notes
- **After upgrading, disable and then re-enable the Reporting API loggers**
- Data collected with 1.2.2610.0 may contain duplicates at slice boundaries and in time ranges where the error repeated
- For high-volume collection, use the S3 logger (see the install guide)
Improvements
- Query validation for genian-edr-search and genian-edr-threats: a query that would pass for "no results" is reported as an error before it runs
- Field names missing from the index or with the wrong case (dsl, stats): similar field names are suggested
- Function notation such as `ip("...")`, `==` or `=` comparisons, lower case `and`/`or`/`not`, quotes around a field name or a whole condition, and wildcards inside quotes
- A non-address value or a wildcard (`10.0.*`) on an IP field: the CIDR notation is suggested
- Syntax errors such as unbalanced parentheses or quotes, a field without a value, or a range without `TO` are reported with their position
- Genian EDR threat analyst template
- Added dsl writing rules (field name case, IP block notation, operators)
- Added how to decode PowerShell encoded commands (-EncodedCommand) with frombase64() and decode(), and allowed the rex command
Bug Fixes
- On browsers older than Chrome 113, screen requests were sent without their parameters and failed with "profile should be not null"
- That error was wrongly shown as a missing profile
New Features
- Query commands that retrieve audit logs, content activities, and content in the trash (`confluence-audit-logs`, `confluence-content-activity`, `confluence-content-trash`)
Improvements
- Page Size option for the audit log, content activity, and content trash loggers
- Slice Minutes option for the audit log and content activity loggers
- Automatic retry on temporary Confluence server errors (HTTP 502, 503, 504)
Changes
- Default read timeout of the `Confluence` connect profile changed from 30 to 60 seconds
Bug Fixes
- Fixed the content activity logger failing with timeouts (Read timed out, HTTP 504) on sites with large or frequently edited documents
Notes
- The AI Assistant panel and the Forensic Investigator template require AI agent app 1.11.2611.0 or later
- Query command error codes are consolidated into the 460xxx range (the former 320xxx and 803xxx codes change)
- The search parameter keyword is renamed to keywords
- Definitions are converted once at startup after the upgrade
- Artifact type categories are linked to the new category catalog, and text with no matching category becomes a custom category
- Unmodified built-in detection rules move to English base values with per-language translations; fields you edited are kept as they are
Improvements
- AI Assistant panel: open the AI agent chat beside any screen and talk while the screen stays in view
- Resizable panel; the panel and conversation persist across screens, and reopening restores the last conversation
- Shows only conversations with the Forensic Investigator and selects it for new chats
- New AI agent template: Forensic Investigator (with 19 skills)
- Investigates from detections through supporting records, per-stage pivots, tags and the timeline
- New query commands
- Scans: forensic-artifacts, forensic-raw-events, forensic-events
- Per evidence type: forensic-reg-logs, forensic-evtx-logs, forensic-mft-logs, forensic-usnjrnl-logs, forensic-web-logs, forensic-text-logs, forensic-pcap-logs
- forensic-schema (artifact schema columns), forensic-artifact-types (artifact type definitions)
- Scope by project (campaign-guids) and host (host-guids); the per-image tables are resolved automatically
- Reworked localization and Japanese support
- Names and descriptions of evidence types, artifact types, detection rules and log schemas are kept as per-language translations
- Edit translations per language on the admin screens; edited and newly created definitions keep their translations
- Event messages and rule names appear in the reader's language (forensic-events locale option)
- Japanese for screens, query command help and output fields, and dates
- New Artifact Categories admin menu: add, edit, delete and order categories
- Tag filter
- Tag color filter on the Logs, File System and Registry screens as well
- Filters every tagged item, no longer capped at 1,000
- Browsing screens
- Reads without a default sort so the first page shows sooner
- Tree clicks, Enter and condition filters search at once; global filter and checkbox changes search after a short pause
- File System includes subfolders by default; host nodes can be selected on File System and Registry
- Clicking the same tree node again searches again
- Shows "No data" when there are no results
Bug fixes
- Installing over a version with changed definitions stopped the artifact catalog setup and left it empty
- Search terms with separators, such as x.exe, found nothing
- Detail panel width was computed from the window and went off when the AI panel was open
- Dropdowns and popups stayed open after clicking into the AI panel or moving focus away
Notes
- The AI Assistant panel and the Genian EDR Threat Analyst template require AI agent app 1.11.2611.0 or later
- The AI Assistant panel appears only when the AI agent app is enabled and the license includes the AI assistant feature
Improvements
- AI Assistant panel: open the AI agent chat beside any screen and talk while the screen stays in view
- Resizable panel; the panel and conversation persist across screens, and reopening restores the last conversation
- Shows only conversations with the Genian EDR Threat Analyst and selects it for new chats
- New AI agent template: Genian EDR Threat Analyst
- Investigates detection evidence, the attack story line, related indicators and affected endpoints, then reports a verdict and recommended actions
- On request and with approval, changes threat states and registers IOCs and YARA rules
- 33 new query commands
- Threats: genian-edr-threats, genian-edr-threat, genian-edr-threat-indicators, genian-edr-threat-devices, genian-edr-threat-history, genian-edr-storyline, genian-edr-set-threat-state
- Endpoints: genian-edr-endpoints
- IOC lookup: genian-edr-malware-hashes, genian-edr-malicious-ips, genian-edr-goodware-hashes, genian-edr-good-ips
- IOC registration and removal (single and batch): genian-edr-add-*, genian-edr-remove-*, *-batch
- YARA rules: genian-edr-yara-rules, genian-edr-validate-yara-rule, genian-edr-add-yara-rule, genian-edr-set-yara-rule-state, genian-edr-remove-yara-rule
- Output field names follow the log schema
- Japanese help for genian-edr-search
- Compatibility with older Genian EDR consoles (3.0.9.0701)
- Detects from console responses whether the Elasticsearch proxy path is encrypted and which threat state API to use
- Better error display: a plain sentence with collapsible details, and a single banner for shared errors on the threat monitoring screen
Bug fixes
- Elasticsearch error reasons were taken from a truncated response body, hiding the cause
New Features
- Added 14 related entity query commands (IP, domain, hash, URL, vulnerability, malware, company)
- `rf-related-{type}`, `rf-related-{type}-batch`
- Added 14 link query commands (IP, domain, hash, URL, email, vulnerability, malware)
- `rf-links-{type}`, `rf-links-{type}-batch`
Improvements
- Added per-VDOM lookup to the `fortigate-firewall-policies`, `fortigate-addresses`, and `fortigate-address-groups` query commands
- Use the `vdom` option to query a specific VDOM only
- Added the `vdom` output field (the VDOM the item belongs to)
- Added the `risk` field (LOW/MEDIUM/HIGH) to the nine log schemas that have a `level` field
Changes
- When the `vdom` option is omitted, the three query commands now query all VDOMs on the device instead of only the API account's default VDOM
- If the VDOM list cannot be retrieved (no permission, or VDOMs not enabled), only the default VDOM is queried as before and the `vdom` field is empty
- Changed the type of the country fields (`src_country`, `dst_country`) in the `fortigate-app-control` and `fortigate-vpn-client` log schemas to `COUNTRY`
Bug Fixes
- Fixed the empty `comment` field in the `fortigate-firewall-policies` query command
Upgrade Notes
- Queries and scheduled queries written for 1.4.2606.1 or earlier with these three commands now return results from all VDOMs; specify the `vdom` option to keep querying a single VDOM
- Querying all VDOMs requires reading the VDOM list (global settings), so an API account bound to a specific VDOM may only get its default VDOM
Initial release
- DDD log parser, log schemas, logger model
- Supports CEF format Syslog only
Feature Enhancements
Added Network Range Lookup Command
- Directly look up an IP address's network range and map it to event streams without pre-specifying a range group.
Improved Ticket Synchronization
- Added detection rule resync capabilities, enabled MITRE ATT&CK technique mapping on duplicated tickets, added dark theme support, and improved overall UI usability.
Initial release
- Logger model and parser for alarm syslog from the Cisco Secure Network Analytics (Stealthwatch) management console (SMC)
Bug Fixes
- Fixed an issue where session-terminating block actions (`reset-both`, `reset-server`, etc.) in threat logs were classified as `DETECT`
- Fixed an issue where some block actions in threat logs, such as `drop` and `sinkhole`, were stored as raw values without normalization
Changes
- Changed threat log `action` normalization: `DETECT` for detected and allowed, `BLOCK` for blocked, `UNKNOWN` for others
- Undefined traffic log `action` values are now normalized to `UNKNOWN`
- API logger logs now also provide the raw value before normalization in the `action_original` field
Log Schema Changes
- Added the `xff_ip` (XFF address) field to the `Palo Alto Networks NGFW 위협` and `Palo Alto Networks NGFW 트래픽` log schemas, and listed the existing `action_original` field
- Added the `xff` (X-Forwarded-For header) field to the `Palo Alto Networks NGFW 위협` log schema
Upgrade Notes
- Threat log `action` values differ from version 1.8.2610.0 and earlier, so queries and detection rules that use these values must be updated
- Use `DETECT` and `BLOCK` instead of `PERMIT`, `DENY`, and `DROP`
- Query `reset-*` actions as `BLOCK` instead of `DETECT`