Docs/Release Notes

Release Notes

Every change to the ANTS Platform ships as a numbered release. This page lists every platform release from v0.1.2 (2025-12-02) to v0.1.265 (2026-10-16), newest first, and the endpoint component history by month.

Summary

Platform versionv0.1.265 (2026-10-16)
Platform releases, last 12 months209
Platform releases, total209
Browser extension version0.1.19
Windows endpoint agent version0.1.18
macOS endpoint agent version0.1.28

Releases are cumulative: each version includes everything before it. Self-hosted and dedicated deployments upgrade on their own schedule using the Helm chart for the chosen version. Planned capabilities are listed on the roadmap.

Platform releases

October 2026

64 releases.

v0.1.265

Released 2026-10-16.

Security

  • Security hardening

v0.1.264

Released 2026-10-16.

Maintenance release: dependency updates and internal improvements.

v0.1.263

Released 2026-10-16.

Features

  • Shadow AI: Ingest and inventory MCP server activity from the macOS agent

Fixes

  • Identity and access: Bound API key cache staleness so revoked keys stop working

v0.1.262

Released 2026-10-16.

Maintenance release: dependency updates and internal improvements.

v0.1.261

Released 2026-10-16.

Fixes

  • AI Gateway: Budget AI UI feedback - sheets for forms, keys page layout

v0.1.260

Released 2026-10-16.

Features

  • AI Gateway: Budget hierarchy insights and model-group trace names

v0.1.259

Released 2026-10-16.

Features

  • Shadow AI: One collector list for Windows and Mac

v0.1.258

Released 2026-10-16.

Maintenance release: dependency updates and internal improvements.

v0.1.257

Released 2026-10-16.

Features

  • Shadow AI: Chart view switcher and Departments cards on AI Usage Intelligence
  • Platform: Duotone icon set for sidebar, header and Usage Intelligence tabs

v0.1.256

Released 2026-10-16.

Maintenance release: dependency updates and internal improvements.

v0.1.255

Released 2026-10-16.

Maintenance release: dependency updates and internal improvements.

v0.1.254

Released 2026-10-14.

Fixes

  • Deployment: Drain web and gateway pods before SIGTERM with a 10s preStop

v0.1.253

Released 2026-10-14.

Maintenance release: dependency updates and internal improvements.

v0.1.252

Released 2026-10-14.

Fixes

  • Identity and access: Make Okta discovery tolerate slow DNS, add web dnsConfig

v0.1.251

Released 2026-10-13.

Fixes

  • Shadow AI: Refuse an unreadable release instead of redirecting to S3

v0.1.250

Released 2026-10-13.

Fixes

  • Validate Origin on tRPC mutations

v0.1.249

Released 2026-10-12.

Features

  • AI Gateway: Per-project health check interval, and report a model as degraded when its traffic fails

v0.1.248

Released 2026-10-12.

Maintenance release: dependency updates and internal improvements.

v0.1.247

Released 2026-10-12.

Fixes

  • Clear high and critical dependency CVEs

v0.1.246

Released 2026-10-10.

Features

  • Shadow AI: Limit what the Windows agent collects (settings + enroll script)

Fixes

  • Platform: Stop the page overflowing by a scrollbar width on every screen

v0.1.245

Released 2026-10-10.

Maintenance release: dependency updates and internal improvements.

v0.1.244

Released 2026-10-10.

Fixes

  • Resolve a customer's own user id from metadata

v0.1.243

Released 2026-10-09.

Maintenance release: dependency updates and internal improvements.

v0.1.242

Released 2026-10-09.

Maintenance release: dependency updates and internal improvements.

v0.1.241

Released 2026-10-09.

Maintenance release: dependency updates and internal improvements.

v0.1.240

Released 2026-10-09.

Maintenance release: dependency updates and internal improvements.

v0.1.239

Released 2026-10-09.

Maintenance release: dependency updates and internal improvements.

v0.1.238

Released 2026-10-09.

Features

  • Platform: ANTS icon set for sidebar, header controls and info markers

v0.1.237

Released 2026-10-08.

Fixes

  • Shadow AI: Point the Graph / Copilot Studio connector at api.antsplatform.ai

v0.1.236

Released 2026-10-08.

Maintenance release: dependency updates and internal improvements.

v0.1.235

Released 2026-10-08.

Maintenance release: dependency updates and internal improvements.

v0.1.234

Released 2026-10-08.

Maintenance release: dependency updates and internal improvements.

v0.1.233

Released 2026-10-08.

Maintenance release: dependency updates and internal improvements.

v0.1.232

Released 2026-10-08.

Maintenance release: dependency updates and internal improvements.

v0.1.231

Released 2026-10-08.

Maintenance release: dependency updates and internal improvements.

v0.1.230

Released 2026-10-08.

Maintenance release: dependency updates and internal improvements.

v0.1.229

Released 2026-10-08.

Maintenance release: dependency updates and internal improvements.

v0.1.228

Released 2026-10-08.

Features

  • Shadow AI: Copilot Studio traces in the Graph connector service (Application Insights)

Fixes

  • Shadow AI: Devices list, detail sheet and user detail agree

v0.1.227

Released 2026-10-08.

Maintenance release: dependency updates and internal improvements.

v0.1.226

Released 2026-10-08.

Fixes

  • Platform: Retire the beta pill and call the product Budget AI in its own screens

v0.1.225

Released 2026-10-08.

Maintenance release: dependency updates and internal improvements.

v0.1.224

Released 2026-10-08.

Fixes

  • Platform: Show the sentence a rejected field produced instead of its Zod issue array

v0.1.223

Released 2026-10-07.

Maintenance release: dependency updates and internal improvements.

v0.1.222

Released 2026-10-07.

Maintenance release: dependency updates and internal improvements.

v0.1.221

Released 2026-10-07.

Maintenance release: dependency updates and internal improvements.

v0.1.220

Released 2026-10-07.

Maintenance release: dependency updates and internal improvements.

v0.1.219

Released 2026-10-07.

Maintenance release: dependency updates and internal improvements.

v0.1.218

Released 2026-10-06.

Maintenance release: dependency updates and internal improvements.

v0.1.217

Released 2026-10-06.

Maintenance release: dependency updates and internal improvements.

v0.1.216

Released 2026-10-06.

Maintenance release: dependency updates and internal improvements.

v0.1.215

Released 2026-10-06.

Fixes

  • Deployment: Give the gateway its own service account so its traces reach the blob bucket
  • AI Gateway: Collapse the key row actions into a menu and move the tag filter beside the rows it filters
  • AI Gateway: Settle the advertised base URL on one form so a copied snippet actually resolves

v0.1.214

Released 2026-10-06.

Maintenance release: dependency updates and internal improvements.

v0.1.213

Released 2026-10-06.

Maintenance release: dependency updates and internal improvements.

v0.1.212

Released 2026-10-05.

Maintenance release: dependency updates and internal improvements.

v0.1.211

Released 2026-10-05.

Features

  • Budget AI (beta version)

v0.1.210

Released 2026-10-05.

Features

  • Tracking of non AI apps

v0.1.209

Released 2026-10-05.

Fixes

  • Clear LLM errors, refresh model lists and defaults

v0.1.208

Released 2026-10-05.

Maintenance release: dependency updates and internal improvements.

v0.1.207

Released 2026-10-02.

Maintenance release: dependency updates and internal improvements.

v0.1.206

Released 2026-10-02.

Features

  • Settings redesign, header switcher, onboarding and setup flow, TrailMap follow-ups

v0.1.205

Released 2026-10-01.

Features

  • One manifest per run, generated from PostgreSQL

v0.1.204

Released 2026-10-01.

Maintenance release: dependency updates and internal improvements.

v0.1.203

Released 2026-10-01.

Fixes

  • Block outdated models, add current ones, explain why

v0.1.202

Released 2026-10-01.

Features

  • Notifications: In-app notifications - bell, RBAC-aware recipients, long-polling
  • A re-export reaches the present and stops touching the cron's cursor
  • Telemetry: Surface dropped/unrecognized log records instead of a misleading empty 200
  • Dd-traces and opentelemetry sdk
  • Make SIEM export stale-lock timeout configurable
  • Show run windows with times and progress in Recent Runs
  • Write a manifest per interval, including empty ones
  • Shadow AI: Fleet token table + device certificate detail UI
  • Record rows exported per dataset, and show all four in Recent Runs
  • TrailMap: Search and filter across diagrams and replays
  • Shadow AI: Gate provider requests and nomination queue behind env

Fixes

  • Shadow AI: Stop closed browsers marking a live device Offline and Revoked
  • Shadow AI: Restore the identity-map sort Prisma accepts
  • Shadow AI: Keep a person's email when their freshest install reports none
  • Telemetry: Honour selected discovered keys and govern SDK metadata.attributes per field
  • Deployment: Give web a writable .next/cache so Next 15 image optimization does not crash pods
  • Clear Wiz Critical/High CVEs in worker, trace-agent and data-store-service images
  • Platform: Drop npm, corepack and yarn from the web runtime image
  • Telemetry: Stamp server-attested provenance on ingested telemetry to bound audit-trail forgery
  • Platform: Build golang-migrate v4.20.1 from source to clear Wiz CVEs and libxml2
  • Run Now returns zero rows on every click
  • Stop an abandoned run blocking a project indefinitely
  • Shadow AI: Alias top-agent subquery columns so ClickHouse resolves them locally
  • Treat null and exact-cased blanks as blank in key lookup
  • TrailMap: Search-and-filter review follow-ups
  • Set an explicit S3 partSize, paired with a queueSize that fits the worker
  • Platform: Explain tRPC errors instead of leaking internals
  • DAST quick wins , HSTS, generic 403, control-char validation
  • Match customer attribute keys case-insensitively
  • TrailMap: Build the node search indexes without CONCURRENTLY
  • Telemetry: Stop key selection dropping metadata keys
  • Stop standardized attributes substituting platform data for customer data
  • Align every export boundary to the interval grid
  • Shadow AI: Restrict registry endpoints to enrollment keys only
  • Stop the standardized export dropping traces and double-counting scores
  • Record resource details and actor identity for SIEM Integration
  • MTLS device-auth reliability and UX fixes
  • Address the code review (#164 follow-up)
  • Apk upgrade in the runner stage to pick up Alpine security patches
  • Safer agent-authentication rollout UX

September 2026

29 releases.

v0.1.184

Released 2026-09-19.

Features

  • Re-export the configured range on demand

Fixes

  • Keep the server barrel out of the settings page bundle
  • User graph range to 90 days, whole-number KPI trends, top AI app / CLI tool

Security

  • Security hardening

v0.1.183

Released 2026-09-18.

Maintenance release: dependency updates and internal improvements.

v0.1.182

Released 2026-09-17.

Maintenance release: dependency updates and internal improvements.

v0.1.181

Released 2026-09-17.

Fixes

  • Scope queue job ids to the run

v0.1.180

Released 2026-09-17.

Fixes

  • Pass web.image.tag to the ClickHouse stage too

v0.1.179

Released 2026-09-17.

Maintenance release: dependency updates and internal improvements.

v0.1.178

Released 2026-09-17.

Maintenance release: dependency updates and internal improvements.

v0.1.177

Released 2026-09-17.

Features

  • Identity and access: Fold ROI AI into Gauge AI

v0.1.176

Released 2026-09-16.

Fixes

  • Telemetry: Fail closed when the key-selection config store is unreachable

v0.1.175

Released 2026-09-15.

Fixes

  • Telemetry: Drop duplicate status_code sample left by the #150/#154 merge

v0.1.174

Released 2026-09-15.

Features

  • Redis result cache for the AI Usage Intelligence tab

v0.1.173

Released 2026-09-15.

Maintenance release: dependency updates and internal improvements.

v0.1.172

Released 2026-09-15.

Fixes

  • Telemetry: Add sample values for the 32 DigitalTwin/Scout keys

v0.1.171

Released 2026-09-15.

Maintenance release: dependency updates and internal improvements.

v0.1.170

Released 2026-09-13.

Fixes

  • Stop custom ranges over 30 days 400ing the activity graph

v0.1.169

Released 2026-09-13.

Features

  • KPI trend cards across every dashboard

v0.1.168

Released 2026-09-12.

Security

  • Security hardening

v0.1.167

Released 2026-09-12.

Fixes

  • AI Command Center: Resolve agent icons consistently across surfaces

v0.1.166

Released 2026-09-11.

Fixes

  • AI Command Center: Stop registry touches masquerading as agent activity

v0.1.165

Released 2026-09-11.

Features

  • OTEL mascot, persona characters, and unified avatar identity

Fixes

  • Make concurrent accepts idempotent

v0.1.164

Released 2026-09-09.

Maintenance release: dependency updates and internal improvements.

v0.1.163

Released 2026-09-08.

Maintenance release: dependency updates and internal improvements.

v0.1.162

Released 2026-09-08.

Features

  • New fields in SIEM integration

Fixes

  • Standardize date formatting, eventType/status, and incoming/outgoing type
  • Identity and access: GitHub SSO fails with OAuthCallback -- drop the client block

v0.1.161

Released 2026-09-04.

Fixes

  • Identity and access: Reject HTML and template syntax in custom role name/description

v0.1.160

Released 2026-09-04.

Features

  • CSP and header hardening

Fixes

  • ApplicationName CrewAI fallbacks + 8 Kore.ai-bridge attributes + resolution-order docs

v0.1.158

Released 2026-09-02.

Maintenance release: dependency updates and internal improvements.

v0.1.157

Released 2026-09-02.

Maintenance release: dependency updates and internal improvements.

v0.1.156

Released 2026-09-02.

Fixes

  • Standardized Export attribute coverage audit + Azure Content-Type + ingestion-queue userId/sessionId backfill
  • Platform: Cap V8 heap and raise the memory limit for blob-storage exports

v0.1.155

Released 2026-09-01.

Maintenance release: dependency updates and internal improvements.

August 2026

62 releases.

v0.1.154

Released 2026-08-31.

Fixes

  • Azure Service Principal, attribute selection, Run Now, and empty-export bugs

v0.1.152

Released 2026-08-31.

Maintenance release: dependency updates and internal improvements.

v0.1.151

Released 2026-08-31.

Maintenance release: dependency updates and internal improvements.

v0.1.150

Released 2026-08-31.

Maintenance release: dependency updates and internal improvements.

v0.1.149

Released 2026-08-31.

Features

  • ADLS/Azure export - Service Principal auth, attribute filtering, Standardized Export

v0.1.148

Released 2026-08-30.

Security

  • Security hardening

v0.1.147

Released 2026-08-28.

Fixes

  • Telemetry: Decouple CrewAI span classification from key-selection content
  • Shadow AI: Correct Firefox gecko ID in enrollment script generator

v0.1.146

Released 2026-08-26.

Maintenance release: dependency updates and internal improvements.

v0.1.145

Released 2026-08-26.

Maintenance release: dependency updates and internal improvements.

v0.1.144

Released 2026-08-26.

Maintenance release: dependency updates and internal improvements.

v0.1.143

Released 2026-08-26.

Maintenance release: dependency updates and internal improvements.

v0.1.142

Released 2026-08-25.

Maintenance release: dependency updates and internal improvements.

v0.1.141

Released 2026-08-24.

Maintenance release: dependency updates and internal improvements.

v0.1.140

Released 2026-08-24.

Features

  • Telemetry: Two-bucket transfer-list picker for key selection

Fixes

  • Stop forcing ML image tags on core-only releases

v0.1.139

Released 2026-08-24.

Maintenance release: dependency updates and internal improvements.

v0.1.138

Released 2026-08-24.

Features

  • Telemetry: Organization-level input/output key selection for OTel ingestion

v0.1.137

Released 2026-08-24.

Features

  • Org-level data retention

v0.1.136

Released 2026-08-23.

Features

  • Integration of directory at org level
  • API ingestion only
  • Added pkce, nonce and state in auth code flow

v0.1.135

Released 2026-08-19.

Maintenance release: dependency updates and internal improvements.

v0.1.134

Released 2026-08-17.

Maintenance release: dependency updates and internal improvements.

v0.1.133

Released 2026-08-17.

Maintenance release: dependency updates and internal improvements.

v0.1.132

Released 2026-08-17.

Features

  • Gate API Keys tab behind NEXT_PUBLIC_HIDE_ORGANIZATION_API_KEYS_SETTINGS

v0.1.131

Released 2026-08-17.

Maintenance release: dependency updates and internal improvements.

v0.1.130

Released 2026-08-14.

Fixes

  • Stop forcing connector image tags on core-only releases

v0.1.129

Released 2026-08-14.

Maintenance release: dependency updates and internal improvements.

v0.1.128

Released 2026-08-13.

Maintenance release: dependency updates and internal improvements.

v0.1.127

Released 2026-08-13.

Features

  • User reconciliation for okta

v0.1.126

Released 2026-08-13.

Maintenance release: dependency updates and internal improvements.

v0.1.125

Released 2026-08-13.

Features

  • Api key with 6 months expiry with notifications for 30, 14, 7 and 1 days

v0.1.124

Released 2026-08-13.

Features

  • CreateCors and hardened the http methods

v0.1.123

Released 2026-08-13.

Fixes

  • Navigation UI - skeleton, data rendering respective to menu and …

v0.1.122

Released 2026-08-13.

Fixes

  • Shadow AI: Remove self-hosted EAP zip install path

v0.1.121

Released 2026-08-12.

Fixes

  • Dependency issues

v0.1.120

Released 2026-08-12.

Maintenance release: dependency updates and internal improvements.

v0.1.119

Released 2026-08-11.

Fixes

  • Added netscope ips
  • Docs update

v0.1.118

Released 2026-08-11.

Maintenance release: dependency updates and internal improvements.

v0.1.117

Released 2026-08-11.

Maintenance release: dependency updates and internal improvements.

v0.1.116

Released 2026-08-11.

Features

  • Exception handling
  • Server side validation for exports
  • Input output masking using ff

Fixes

  • Malformed metadata
  • Cache invalidation
  • The hook fails open, and the doc claims the opposite

v0.1.115

Released 2026-08-11.

Maintenance release: dependency updates and internal improvements.

v0.1.114

Released 2026-08-11.

Maintenance release: dependency updates and internal improvements.

v0.1.113

Released 2026-08-11.

Features

  • V2 residual OTEL reconciliation for the Users export
  • Owner data exports for Shadow AI Users/Agents, with size + window caps

Fixes

  • Address PR #40 review findings , missing down.sql, DATABASE_URL YAML quoting, stale session-age validator
  • Address pre-merge audit findings , preset-range export bug, unrelated docker-compose diff
  • Correct PR-flagged issues , global size cap docs, drop unrelated .gitattributes, honor on-screen range for Users export

v0.1.112

Released 2026-08-11.

Maintenance release: dependency updates and internal improvements.

v0.1.111

Released 2026-08-11.

Maintenance release: dependency updates and internal improvements.

v0.1.110

Released 2026-08-10.

Maintenance release: dependency updates and internal improvements.

v0.1.109

Released 2026-08-07.

Maintenance release: dependency updates and internal improvements.

v0.1.108

Released 2026-08-07.

Fixes

  • Identity and access: Eliminate remaining session flicker root causes
  • Suppress credit exhaustion modal for Enterprise plan orgs

v0.1.107

Released 2026-08-07.

Maintenance release: dependency updates and internal improvements.

v0.1.106

Released 2026-08-07.

Features

  • SIEM integration, with s3 scheduled audit logs hourly

Fixes

  • Replace raw session.status comparisons with isInitialSessionLoad/hasAuthenticatedSession to prevent flicker on background refresh

v0.1.105

Released 2026-08-06.

Maintenance release: dependency updates and internal improvements.

v0.1.104

Released 2026-08-06.

Fixes

  • Disable recommendationEngine + piiPciScheduler; trim build-core to 4 images

v0.1.103

Released 2026-08-06.

Maintenance release: dependency updates and internal improvements.

v0.1.102

Released 2026-08-06.

Maintenance release: dependency updates and internal improvements.

v0.1.101

Released 2026-08-05.

Maintenance release: dependency updates and internal improvements.

v0.1.99

Released 2026-08-05.

Maintenance release: dependency updates and internal improvements.

v0.1.98

Released 2026-08-05.

Features

  • Coding agent

Fixes

  • Serialization
  • Bake NEXT_PUBLIC_HIDE_* flags into web image at build time

v0.1.97

Released 2026-08-05.

Features

  • Surface auth events, add page-access logging, and audit export completion
  • Extend schema for auth event auditing

Fixes

  • Populate orgId on auth events via per-org fan-out
  • Guard logAccess per-page, share complete IP extraction
  • Stop double-logging user.login, fix userId resolution

v0.1.96

Released 2026-08-05.

Features

  • Hide Protected Prompt Labels settings tab

Fixes

  • Import z from zod/v4 instead of zod v3
  • Align useCurrentOrganization priority with usePlan + zod/v4 import
  • Close gaps in Enterprise billing hide (follow-up to #26)

v0.1.95

Released 2026-08-05.

Fixes

  • Exclude input/output from traces-list response by default

v0.1.94

Released 2026-08-05.

Fixes

  • Exclude model_name and finops_job_configuration_id from outer WHERE on both trace paths

v0.1.93

Released 2026-08-05.

Fixes

  • UX updates and user access removal

v0.1.92

Released 2026-08-04.

Features

  • Hide billing for Enterprise orgs + hide Scores/Evaluation tab

v0.1.91

Released 2026-08-03.

Features

  • Identity and access: Sign-in code step and per-organization MFA toggle
  • Custom roles
  • Modular RBAC
  • Added okta provider
  • Okta sso config
  • Shadow AI: Delete the Browsers filter dimension; make OS per-tab (D34)
  • Shadow AI: Scope filters per surface , AI Tools tabs (D31) + Usage Intelligence honors category/department end-to-end (D32)
  • Org-level feature flags for Guardian AI visibility (AI Guardrails / AI Violations)).
  • Shadow AI: Push users-page filters end-to-end across Guardian queries
  • Shadow AI: AI Agents tab standalone chrome + delete sanction/risk filter dimension
  • Shadow AI: Guardian telemetry chips on Devices page via email bridge (D26 range-log hardening included)
  • Shadow AI: Never-throw OTEL model arm , log + degrade to endpoint models / org-wide scope on failure (D20)
  • Shadow AI: Activity graph legibility , adaptive fan-out, combined endpoint+OTEL counts on nodes, agent model breakdown via observations
  • Shadow AI: Unify AI-prefixed persona taxonomy and user KPI tiles
  • Shadow AI: Integrate Guardian OTEL metrics
  • Shadow AI: Reuse Command Center TimeRangeSelector and add source badges to tables for consistency
  • Align Total Agents tile to windowed view , active agents per range with all-time registry context
  • Shadow AI: Structured logging and sanitized errors for usage-intelligence and users/detail endpoints
  • Shadow AI: AI coverage KPI (monitored share of active AI users) replaces Activation rate
  • Shadow AI: Include OTEL-only users in Active AI Users and fix activation rate
  • Shadow AI: Add "AI Agents" category for OTEL agents
  • Shadow AI: Combined endpoint + Guardian totals on the user detail page
  • Shadow AI: Combined endpoint + Guardian agent totals in Users and Departments tables
  • Shadow AI: Coalesce OTEL sessions and interactions into every org surface
  • Shadow AI: Identity-scope OTEL usage intelligence, drop the agent allowlist
  • Shadow AI: Per-user Guardian OTEL rollup and OTEL-only rows in users summary
  • Shadow AI: Show Guardian OTEL users in the Users list and detail pages
  • Shadow AI: Add Top AI agents by users Guardian OTEL chart to Usage Intelligence tab
  • Shadow AI: Show all of a user's agents in the activity card, not just coding agents
  • Shadow AI: AI Usage Intelligence tab (PR #15)
  • Shadow AI: Render per-department category mix as a chip cell
  • Shadow AI: Add per-department Usage by category breakdown
  • Replace AI adoption page with AI usage intelligence tab
  • Shadow AI: Wire Q9/sanction changes into orchestrator, add tool logo markers and partial-data banner
  • Shadow AI: Wire up AI Usage Intelligence tab and tRPC router
  • Shadow AI: Implement AI Usage Intelligence tab (server + UI)
  • UX enhancements of guardrails
  • Shadow AI: Add OTEL agent_name mapping and export enrichUsersWithIdentity
  • Email changes
  • Reorder sidebar groups, hide prompts/evals, move tracing to Guardian AI
  • Search bars, filters and auto select after creating an agents
  • Shadow AI: Add 'uia-heuristic' detection kind for Windows UI Automation
  • Shadow AI: Add MetricLine component for improved metrics display
  • Guardrails: Qwen-only in-cluster GPU backend; remove local Llama/DuoGuard models
  • Telemetry: Copilot cost + real model (github.copilot.cost, response model)
  • Tagging changes for online mode, showing online voilations in ants UI
  • Telemetry: Auto-attribute Codex + Copilot traces to a user
  • Telemetry: Consistent coding-agent parsers + one-prompt-per-trace
  • Telemetry: Dev-only raw OTLP capture tap
  • Shadow AI: Persona + user filter on Users page with single-source charts
  • Offline Guardrails 1. Changed Service names 2. guardrails UI creation redesign 3. model change and logic change
  • Telemetry: Port net-new CLI-agent changes onto develop
  • Shadow AI: Add screen-local user filter to Users page
  • Update image handling for Copilot SVG and adjust styles
  • Model swapping for online mode
  • Guardrails: Remote Qwen3Guard offline detection + offline/online policy config UI
  • Sync model catalog from upstream langfuse + track [1m] overlay
  • Add Zscaler ICAP favicon support
  • Shadow AI: Add Uploads/Tools and Monitored By columns to user Recent Activity
  • Shadow AI: Implement drag-and-drop sorting for navigation items and groups
  • Shadow AI: Versioned release filenames + unified device capability tiles
  • Shadow AI: Update device node height and enhance provider lookup functionality
  • Shadow AI: Add project folder support to metrics and UI components
  • Shadow AI: Enhance data handling and reporting
  • Shadow AI: Add 'kind' property to Windows tool configurations
  • Shadow AI: Tabbed apps UI, parser session table, Windows agent settings & identity fix
  • Shadow AI: Add Microsoft 365 Copilot + promote consumer Copilot to intercept
  • Telemetry: CrewAI Enterprise + generic gen_ai semconv parser
  • Add Windows detection for ChatGPT, Microsoft Copilot, GitHub Copilot CLI, Notion, Grammarly
  • Add Aider CLI Windows detection rules
  • Shadow AI: Unify desktop + browser deployment UI behind segmented toggles
  • Shadow AI: Enrich CLI sessions with parser data and file upload tracking
  • Shadow AI: Enhance CLI support and improve uninstall script for Windows agent
  • Shadow AI: Identity v2 cutover + UX cleanup across registry, users, devices
  • Shadow AI: Windows agent ingestion, device dedup, registry fixes
  • Shadow AI: Accept windows-agent as a valid client kind
  • Shadow AI: Internal Apps in extension config + piggyback refresh
  • Shadow AI: Icon + Powered by X plumbing across UI
  • Shadow AI: Registry rebrand , Internal Apps + My Nominations tabs
  • Shadow AI: Internal Apps CRUD + nomination dedup + icon upload tRPC
  • Shadow AI: Icon storage + upload pipeline
  • Shadow AI: Layer 2 Prisma guard + RLS GUC context
  • Shadow AI: Model breakdown + activity graph views
  • Shadow AI: Enhance enrollment script with enroll/unenroll modes and improve client count logic
  • Adopt langfuse pricing tiers + refresh model catalog
  • AI Command Center: Add Inactive status filter and demote inactive agents
  • Shadow AI: Deployment & Keys rework , shared enroll-script Step 0 + inline MDM
  • Shadow AI: Add post-enrollment redirect to Shadow AI dashboard
  • Shadow AI: Per-session active time from hostFlow + session column
  • Shadow AI: Use vendorWebsite as favicon fallback when apiHosts is absent
  • Shadow AI: Include CLI session data across Overview + User Detail aggregations
  • Rebrand plans to Agentic Observability & Guardian AI
  • Shadow AI: Gate pages behind enterprise + refine focus-time calc
  • Shadow AI: Add Apple logo icon component and update macOS section icon
  • Shadow AI: Unify browser/macOS sessions across app & user views
  • Shadow AI: Success panel on enroll page + hostname normalization
  • Shadow AI: Add SourceBadge component and update app detail source tracking
  • Shadow AI: Browser/macOS read-layer unification + registry UX cleanup
  • Enhance risk level handling in shadow AI router
  • Changed signin flow for extension and agent
  • Implement collapsible sections for navigation groups and update route visibility fix: remove unused cost projections from AgentDetailPanel and PerformanceTab components style: add collapsible animations to Tailwind configuration
  • Shadow AI: Add settings overview, deployment, and policy components
  • Shadow AI: Add per-user CLI and browser activity endpoints; enrich provider data with sanctions and categories
  • Shadow AI: Implement CLI tool detail sheet and integrate with CLI tools section
  • Remove legacy Shadow AI activity and session models, and related API endpoints
  • Shadow AI: Add device/file detail sheets, cli tools section and wire new components into pages
  • Shadow AI: Add shared badges, section primitives and risk/confidence constants
  • Shadow AI: Add clients and registry public APIs, update events ingestion
  • Shadow AI: Add clickhouse read/write services, banned key guard and extend router
  • Shadow AI: Implement custom app management with creation, editing, and listing functionalities
  • Shadow AI: Enhance Shadow AI components with sanction management and app detail features
  • Shadow AI: Implement Shadow AI app detail sheet and enhance related components with sanction management
  • Shadow AI: Add app sanction posture (constants, badge, card)
  • Shadow AI: Enhance layout and navigation for Shadow AI features, including new routes and improved component integration
  • Shadow AI: Add apps, devices, reports, users pages integrate into layout, routes, and command menu
  • Shadow AI: Update feature tables, overview, reports, user detail
  • Shadow AI: Add Shadow AI components and shared utilities
  • Add Shadow AI favicon proxy with S3 caching
  • Implement utility functions for Shadow AI including collapseTopN, formatting, debouncing, and user display metadatafeat: implement utility functions for Shadow AI including collapseTopN, formatting, debouncing, and user display metadata
  • Add constants for shadow AI features including app providers, categories, device health, chart styles, and time ranges
  • Add type definitions for shadow AI features including activity, app, device, session, user, and overview
  • Added apis for shadow ai
  • Enhance privacy notice links and Update sorting options to prioritize 'last_active'
  • Implement integration action components for Jira, Slack, and Linear in FinOpsRecommendationsWithSavings
  • Remove VERSION display handling and update font constants to include Poppins
  • Add new integration documentation links for CrewAI, Slack, Jira, and Linear
  • Update documentation links in AnnotationQueuesOnboarding and DatasetsOnboarding components
  • Add isTableAction prop to DeleteEvalConfigButton in EvaluatorTable and FinOpsJobConfigurationDetails
  • Integrate Linear issue creation functionality in FinOps recommendations
  • Improve bot icon implement agent tags context and update BotIcon to use agentId for platform resolution.
  • Add SIL Open Font License text for font software compliance
  • Add font switching feature and optimize image caching
  • Add VERSION file support and improve layout spacing in components
  • Add Linear integration to AI Command Center agent detail
  • Add Linear notification support for compliance, detection, and governance
  • Add Linear support to agent project mappings with 3-way merge
  • Add Linear settings page, connection card, and logo assets
  • Add Linear tRPC router, OAuth handler, and API route
  • Implementation of guadrail feature
  • Enhance governance compliance tracking and UI components with evidence coverage and attestation status
  • Enhance Framework Compliance Summary and Dashboard with Agent Compliance Metrics
  • Enhance risk dashboard and governance components
  • Enhance GovernanceProfileService to track requirement completion and agent compliance details
  • Enhance FrameworkComplianceCard with donut chart and tooltips for compliance status
  • Update risk dimension descriptions and tooltips for clarity and accuracy
  • Implement compliance record management in GovernanceEvidenceService and sync with GovernanceProfileService
  • Add expand/collapse functionality to ConfigurationSummary and PolicyConfigForm components
  • Enhance RiskDashboardPage and RiskSummaryCards with improved navigation and link functionality
  • Add drift alert incident handling for PagerDuty and ServiceNow
  • Add CellTooltip component and integrate it into DriftAlertsTable and DriftNotificationButton for improved tooltip functionality
  • Enhance UI of governance components with improved styling and action slots
  • Add BotIcon component to various governance pages and enhance UI elements
  • Enhance drift alerts functionality and UI across governance sections
  • Enhance drift alerts functionality with pagination, filtering, and improved skeleton UI
  • Add Drift Alerts components and utilities for governance monitoring
  • Add Cost Performance Metrics and Latency & Analytics pages with routing updates
  • Add guardrail policy configuration to agent detail and overview tabs
  • Add ServiceNow/PagerDuty integration badges, fix risk score consistency, and UI improvements
  • Add guardrail violations and drift alerts filters, update status to risk-score based
  • Implement composite risk score and unified risk indicators for AI Command Center
  • Fix build errors update time range options for risk dashboard and drift alerts
  • Enhance compliance violation metrics and drift alert queries with agent filtering
  • Enhance risk and compliance dashboard with agent name handling, improved navigation, and performance optimizations
  • Enhance risk dashboard components with project ID handling, improved drift alert labeling, and performance optimizations in data queries
  • Enhance risk compliance dashboard with project ID handling, error management, and compliance violation schema
  • Add Risk Dashboard components and hooks
  • Implement risk dashboard feature with tRPC router, computation service, and validation schemas
  • Implement risk compliance dashboard with heatmap and drift alerts overview
  • Improve RBAC UI members page in org and project settings panel
  • Add LLM Inventory, My Requests, and Violations previews
  • Enhance ethics bias entity handling and validation in PiiPciTracesTable
  • Production-ready connectors with graceful shutdown, helm improvements, and v0.1.48
  • Add scripts and configuration for managing enterprise plan entitlements
  • Wrap AI compliance and command center components with EnterpriseFeatureGate for plan restrictions
  • Implement enterprise features and entitlement checks
  • Update favicon and icon assets for improved branding
  • Enhance batch processing and ClickHouse integration
  • Update SignIn component to support dark mode for AgenticAnts logo
  • Update AgenticAnts logo with light and dark mode support
  • Low code now code connector
  • Improve compliance section ui and enhance chart color semantics for better data visualization
  • Implement batch message schema and publishing for DuoGuard detection
  • Add driftAlert to AuditableResource type
  • Improve ethics/bias detection accuracy with model-based bias detection
  • Strengthen governance and drift management with runtime validation, automated policy activation, and integrated alerting
  • Improve AI Policy UI consistency and loading states
  • Added llmn trace connector for bedrock and vertex
  • Added credit consumption for ai policies
  • Added credit consumption logic for ai policies
  • AI Governance: Enhance framework compliance data retrieval and UI with filtering options
  • AI Governance: Add Export & Audit endpoints for compliance reports and audit logs - Introduced listReports and getAuditLog endpoints in the AI Governance router. - Implemented corresponding methods in the GovernanceProfileService to handle report listing and audit log retrieval with filtering and pagination. - Added validation schemas for the new endpoints: ListReportsInputSchema and GetAuditLogInputSchema. - Updated the governance router documentation to reflect the new endpoints and their functionalities.
  • AI Governance: Implement dashboard statistics and recent activity endpoints with API integration
  • AI Governance: Add notification fields and feedback mechanisms for Slack and Jira integration
  • AI Governance: Enhance agent mapping integration for Slack and Jira in governance components
  • AI Governance: Enhance governance schemas and database structure with new evidence form fields and agent listing functionality
  • AI Governance: Pass agentName to RuntimeValidationSection for trace navigation
  • AI Governance: Enhance framework compliance mapping and add latest attestation logic
  • AI Command Center: Multi-select compliance notification modal, icon/tooltip fixes, and ServiceNow/PagerDuty removal
  • AI Governance: Update compliance framework identifiers to uppercase
  • AI Governance: Add schema + governance utilities
  • Add PagerDuty incidents page with pagination and agent/status filters
  • Add PagerDuty integration for compliance alerting
  • Add Role-Based Access Control (RBAC) documentation for AI Governance feature
  • Add Role-Based Access Control (RBAC) documentation and update API access details in AI Compliance docs
  • Update request status handling and UI components in AvailableRequestsPage and MyRequestsPage
  • Update MyRequestsPage to clarify status labeling and format date display
  • Enhance ExportAuditPage UI with improved tab styling
  • Add edit functionality and pagination to Drift Alerts section
  • Remove unused activity filters and related components from GovernanceDashboardPage
  • Added servicenow integration
  • Add filter removal functionality to PopoverFilterBuilder and InventoryFilters
  • Remove VIEWER role from AI Compliance API specification
  • Add script to import new models from JSON and update Excel inventory
  • Enhance Evidence Collection and AI Compliance tabs based on user roles
  • Add ethics_bias_labels column to pii_pci_config table
  • Configure ethics-bias-detector with GPU acceleration
  • Add Request Changes and Request Creation modals for AI Governance
  • Add comprehensive API specification doc for AI Compliance feature
  • Implement assignment route for governance profiles and update navigation paths
  • Implement recent activity tracking with filters and detailed activity types
  • Enhance framework compliance section with additional requirements and framework badges
  • Adjust column widths and add Request column in Agent Evidence Table
  • Add Slack and Jira integration buttons to Agent Evidence Table
  • Add components for governance tracking and compliance management - Implemented CoverageProgressBar for visualizing evidence coverage. - Created DriftAlertBadge to display drift alert counts with severity styling. - Developed FrameworkComplianceCard to show compliance status for frameworks. - Added GovernanceEmptyState for displaying when no governance data is available. - Introduced GovernanceLoadingState for loading skeletons in governance components. - Created GovernanceStatCard for reusable stat cards in the Agent Governance dashboard. - Added RiskLevelBadge to display risk classification levels. - Re-exported shared components for easier imports in the governance module. - Defined constants for evidence categories, compliance frameworks, risk levels, and drift types. - Established types for evidence items, governance profiles, drift alerts, and dashboard statistics. - Implemented navigation utilities for AI compliance tabs. - Created pages for managing agent evidence, compliance reports, and audit exports.
  • Add ethics & bias detection with Llama Guard 3 1B
  • Add PostgreSQL readonly user setup script for local development
  • Add ClickHouse dictionary setup guide and PostgreSQL readonly user script
  • Enhance compliance monitoring with improved security and validation
  • Enhance agent list API with compliance data and notifications
  • Telemetry: Add organizationId to trace metadata for compliance monitoring
  • Improve UI consistency and add tooltips for LLM and framework inventory
  • Telemetry: Add organizationId to ingestion events and processing logic
  • Simplify approval UX and unify inventory UI
  • End-to-end compliance support across backend and AI Command Center
  • Enhance ApprovalCell with toggle switch and improve MultiSelectDropdownFilter with search functionality
  • Add phi detection
  • PHI implementation
  • Implement model processing status updates for FinOps evaluations and trace generation
  • Enabled gliner spacy
  • Enhance UX/UI with accessibility and visual improvements
  • Add Framework Inventory UI with dual-tab interface
  • Add Framework Inventory backend and database schema
  • Enhance provider requirements section with Azure deployment details
  • Remove FinOps recommendation toggle and adjust related form fields
  • Add description field to EvalTemplate and update related components
  • Enhance model selection with pricing data and sorting functionality
  • Add LLM Inventory page and filters, enhance routing and table types, Small supporting changes
  • Add frontend UI components and hooks
  • Publish FinOps eval completion to recommendation-engine for failed models
  • Add failed trace count tracking to batch processing and completion messages
  • Enhance batch job status handling with success and error counts for Azure and OpenAI services
  • Implement backend API and business logic
  • Enhance delete confirmation dialog with structured layout and improved accessibility
  • Implement FinOps completion check and publish mechanism in evalQueue
  • Implement deduplication and model normalization for batch processing in RabbitMQ and enhance FinOps completion checks
  • Add local hybrid development setup with ephemeral bastion
  • Add /pr-audit command for automated pre-PR code review
  • FinOps: Add incremental progress tracking for LLM optimizer jobs
  • FinOps: Add UI components for granular job status tracking
  • FinOps: Implement status updates in trace-agent and worker
  • FinOps: Add tRPC endpoints for granular job status tracking
  • FinOps: Add granular status tracking for LLM optimizer jobs
  • AI Command Center: Implement dynamic theme system with variant-based styling
  • AI Command Center: Add themes with CSS variables
  • Upgrade ClickHouse image to version 25.10 for stability of Refreshable Materialized Views
  • Enhance logging infrastructure with Winston logger across services
  • FinOps: Add compact currency formatting for savings display
  • FinOps: Disable comparison models when 3 are selected
  • Add GIN index on tags for Agent model
  • Add script to sync agent tags from ClickHouse to PostgreSQL
  • AI Command Center: Add tag and integration filtering system
  • Add tag and integration filtering to agent APIs
  • Add tag extraction and JSONB support for agent metadata
  • AI Command Center: Enhance project metrics retrieval and filter handling with improved UI updates
  • AI Command Center: Implement dynamic cost filtering with adaptive boundaries and enhanced UI options
  • AI Command Center: Enhance UI components with improved styling, tooltips, and new features
  • AI Command Center: Enhance time range handling and metrics retrieval across agent detail, list, and project metrics APIs
  • AI Command Center: Add cost filter boundaries to project metrics for enhanced UI filtering
  • AI Command Center: Add time range support to project metrics and agent detail APIs
  • AI Command Center: Enhance agent detail and list hooks with time range support
  • AI Command Center: Add agent fields and optimize table TTLs for improved trace analytics
  • AI Command Center: Improve dashboard components
  • AI Command Center: Enhance types and constants for metrics
  • Add scrollbar utilities for improved UI experience
  • Add API endpoints for agent detail, agent list, and project metrics in AI Command Center
  • Implement hooks for agent list, agent detail, and project metrics in AI Command Center
  • Add AgentCard, AgentCatalog, AgentDetailPanel, HeroMetrics, and AgentDiscoveryBar components
  • Add skeleton components for AgentCard, AgentDetail, AgentList, and MetricCard in AI Command Center
  • Add shared components for AI Command Center including FilterDropdown, SearchInput, SearchableFilterDropdown, SortControls, StatCard, SimpleSparkline, TimeRangeSelector, and ViewToggle
  • Add AI Command Center constants, types, and utility functions
  • Complete Azure OpenAI Batch API implementation for FinOps and LLM-as-Judge
  • Implement Azure OpenAI real-time API support for FinOps and LLM-as-Judge flows
  • Add hyphenated trace attributes to OtelIngestionProcessor
  • FinOps: Add projected savings with configurable assumption
  • Implement agent upsert logic for trace and observation events in IngestionService
  • Synchronize agent display name across agents and agent_project_mappings tables
  • Implement agents table for SDK-created agents and update related queries
  • Enhance agent model with mutable display name and update ingestion processor logic
  • Add SSL certificate mounts to web, worker, and recommendation-engine
  • Implement PostgreSQL SSL certificate verification for trace-agent and trace-cron
  • Improve formatting and logging for agent_id generation and validation in ingestion services
  • Add agent_id and agent_name fields for trace and observation processingand update the ingestion code with new implementation
  • Batch inference for llm as a judge
  • Enhance logging for agentId and agentName extraction in OtelIngestionProcessor and IngestionService
  • Add agent_id and agent_name fields for enhanced tracking in observations
  • FinOps: Pass trace count to FinOps job based on sampling
  • Display sampled trace count on sampling slider
  • Add trace-agent-cron deployment and update image tags
  • Feat: Add targetModel parameter to handleBatchActionJob and finopsComparisonCompletionQueueProcessor for trace evaluation
  • Added BATCH_API implementation for the OpenAI in Llm-as-judge
  • Added gemini studio and gemini vertex batch inference
  • Add Vertex AI batch inference support with GCS validation
  • Add OpenAI Batch API support with cron polling service
  • Add batch inference validation with ConfigMap support
  • Add batch inference toggle with cost-savings UI in model selection
  • Add AWS Bedrock inference profile support for batch inference
  • Add BedrockBatchInferenceConfig database model
  • Add Bedrock model validation for batch inference
  • Add AWS Bedrock batch inference scripts and policies
  • Add Bedrock batch inference support with S3 permissions validation
  • Implement real-time credit consumption with Redis distributed locking to prevent deletion refunds
  • Add FinOps job configuration cleanup and delete action in agent details
  • Enhance UI/UX with FinOps improvements and credits breakdown
  • Add delete functionality for FinOps job configurations
  • Redirect to FinOps job detail page after evaluator creation
  • Enforce default evaluation model requirement with improved UX
  • Implement git tag-based version management for releases
  • Add Java SDK integration to quickstart examples
  • Enhance Slack integration with message tracking and UI improvements
  • Hide integration setup links for AI roles and add build documentation
  • Add configurable API host for SDK quickstart examples
  • Implement FinOps recommendations credit tracking and ClickHouse filtering improvements
  • Update image tag and enhance ingress configuration for app and API
  • Implement dual-domain ingress architecture for app and API separation
  • Add automated build system for trace-worker container
  • Optimize Helm charts for cost and production readiness
  • Add recommendation-engine support to Makefiles and Helm charts
  • Add runtime configuration and improve trial user experience
  • Allow trial users to access platform without payment method
  • Migrate K8s deployment to Helm with KEDA autoscaling
  • Add subscription scheduled for cancellation email and reorganize docs
  • Redesign email templates and update billing system
  • Remove overage limit display from CreditDashboard component
  • Identity and access: Implement email verification flow with templates and handlers
  • Implement payment method check for OWNER users during onboarding
  • Recommendation engine setup
  • Streamline onboarding, billing reset, and credit metering
  • Added finops in llm as a judge code
  • Add SSO configuration and environment variable support
  • Integrate RabbitMQ, Implement FinOps agent and job configuration storage
  • Improve integration connection indicators
  • Finops recommendation UI
  • Implemented frontend for finops recommendation
  • Add Slack and Jira integrations with agent team mappings second version
  • Add Slack and Jira integrations with agent team mappings
  • Add AgenticAnts Docker Compose configuration
  • Add comprehensive documentation and FinOps features
  • Initial commit - AgenticAnts AI Platform fork of Langfuse

Improvements

  • Implement parallel Docker builds in Makefile for 5x speedup

Fixes

  • Strip model_name/finops filters from all AMT newExecution paths
  • Goverance ai
  • Rbac for pii-read fix
  • Identity and access: Rate-limit reset codes, refuse SSO accounts
  • Updated rbac and custom roles
  • Custom rbac authorization
  • Custom rbac
  • Feature-toggle race + omitted-flag reset bugs
  • Shadow AI: AI tool landscape zero-value clutter + collision overflow
  • Infinite rendering, setting scoping
  • Shadow AI: Drop two artifacts left by the master merge
  • Shadow AI: Parse ClickHouse timestamps as UTC, not local time (D36)
  • Shadow AI: Apps-page drill-downs honor custom time range (D26)
  • Shadow AI: Custom range on graph drill-downs (D26); getModelBreakdown counts runs + OTEL model authority (D29)
  • Shadow AI: Devices page honors custom time range (D26/D28)
  • Shadow AI: Files page honors custom time range end-to-end (D26)
  • Shadow AI: Custom [from, to] range honored end-to-end on User Detail (D26)
  • Shadow AI: Device detail sheet honors the page time range (was hardcoded 7d)
  • Shadow AI: User-detail heatmap paints Guardian runs (D25)
  • Shadow AI: Count only genuine sessions across usage intelligence
  • Shadow AI: Default unreviewed apps to UNSANCTIONED, normalize provider keys, push economics coalesce into ClickHouse
  • Shadow AI: Show time-range filter inline on User Detail page
  • Self-healing deploy pipeline , no more manual intervention needed
  • Disable qwen-vllm in workloads stage; remove stale image tag set
  • Write both Firefox gecko IDs, Edge machine-level plist, agent config.plist
  • Shadow AI: Add duration_sec_dedup to top-users subquery
  • Shadow AI: Count duration-only CLI sessions in tool detail view
  • Shadow AI: Fix parser 2x double-count bug + add GUI file upload tracking
  • Shadow AI: Update prompt detection for Aider CLI to use app-parser with parser name
  • Shadow AI: Add icons for Grammarly, Notion AI, and Aider
  • Shadow AI: Improve detail display logic for average duration and upload size in app detail view
  • Shadow AI: Enhance prompt detection logic for VS Code Copilot and improve project folder handling in app detail reads
  • Shadow AI: Add onLightBackground prop to AppLogo for better logo visibility on light tiles show monochrome GitHub Copilot logo on light sanction chips
  • Shadow AI: Add optional detectedAt field for accurate file detection timestamps
  • Shadow AI: Enhance app activity metrics to include tool call counts
  • Shadow AI: Enhance session counting to include file uploads in metrics
  • Shadow AI: Update Windows Copilot detection logic and user data paths
  • Including custom block message to sdk response for online mode
  • Telemetry: Don't let resource user.email override a tool's native login
  • Graceful empty state for deleted/expired traces
  • Merge conflict for flate
  • Shadow AI: Provision macOS agent via /usr/bin/security, not the bundled CLI
  • Simplify duration display logic
  • Deployment: Pin data-store-service image via subchart-name key (drop --set workaround)
  • Shadow AI: Edge store install is live (v0.1.12) , drop stale 'pending review' copy
  • Shadow AI: ReturnTo allowlist accepts published Chrome+Edge store IDs
  • Annotate azure resp to break TS7022 inference cycle
  • Shadow AI: Copilot alias + conditional file-events rendering
  • Shadow AI: Consistent parser JOIN across all remaining detail pages
  • Shadow AI: Add missing uploads field to app-detail recentSessions type
  • Shadow AI: Update Mistral domain to correct URL
  • Shadow AI: Tab persistence, detail uploads, detection method, agent liveness
  • Shadow AI: Consistent interaction/session counts across all dashboard surfaces
  • Shadow AI: Complete desktop AX promptDetection spec for macOS agent
  • Shadow AI: Drop Bing from Copilot registry; count token-only CLI sessions
  • Shadow AI: Dedup recent activity rows, strip Parser prefix, fix empty context fallback
  • Shadow AI: Simplify user detail, fix nav collapse, fix settings routing
  • Shadow AI: Parser-only CLI session counts + correct tool/token display
  • Shadow AI: Unify interaction counting across all dashboard views
  • Deployment: Purge stale subchart .tgz archives + preflight to prevent recurrence
  • Shadow AI: Prevent close icon overlapping Open in new tab button in file events sheet
  • Shadow AI: Resolve missing logos for GitHub Copilot CLI and Microsoft 365 Copilot
  • Telemetry: Tag CrewAI agents with framework=crewai so the UI shows the logo
  • Repair macOS aider classifier + split gemini_cli from gemini web row
  • Shadow AI: Migrate Edge extension to Microsoft Edge Add-ons store
  • Shadow AI: Resolve code review issues on CLI parser enrichment
  • Remove windows binaryTarget to unblock Docker builds
  • Reduce web build Node.js heap from 8192 to 6144 MB to prevent OOM
  • Shadow AI: Replace broken Mode tab toggle in WindowsSection with PlatformMdmDisclosure collapsible
  • Shadow AI: Code review fixes , registry admins, icon upload procedure, custom app extended fields
  • Shadow AI: Hotfix-6 , Chrome extension migrated from S3 to Chrome Web Store
  • Shadow AI: Remove redundant overflow styles in ShadowAiUserDetail component
  • Shadow AI: Serve enroll/unenroll scripts as text/plain to bypass Chrome download block
  • Shadow AI: Flush cfprefsd in enroll + unenroll scripts
  • Shadow AI: V0.1.67 hotfix-1 , daily-chart TZ, unenroll completeness, guardrail probe
  • Shadow AI: Windows agent support across UI + data correctness icon + provider display names
  • Shadow AI: Split appHost/apiHost in enrollment allowlist + installer
  • Shadow AI: Correct sidebar active state and browser title on nested pages
  • Pre-PR review findings (gzip bomb, ESO optional fields, tier upsert)
  • Use kebab-case subchart names for helm --set image.tag
  • Create default PricingTier when upserting flat prices
  • AI Command Center: Distinguish sub-tenth-of-a-cent costs from zero in formatCurrency
  • Shadow AI: Align Session Timeline device grouping with Devices page
  • Shadow AI: Debounce favicon loads + clip user-detail x-overflow
  • Shadow AI: Merge per-person across installs, fix file uplaod issue, show checkin history for browser devices
  • Shadow AI: Scope user-detail Files drill-down to the clicked user
  • Shadow AI: Correct MDM plist key prefix + add dashboard indexes
  • Adjust column size and styling for Drift Alerts Table and Notification Button
  • Updated ai registry settings
  • Updated schema for the browser tracking
  • Linear integration missing schema module and enhance guardrail policies
  • Build errors.
  • Open external agenticants.ai links in new tab
  • Make pagerduty_incidents.integration_id nullable to support onDelete: SetNull
  • Update spacy from yanked 3.8.4 to 3.8.14 in guardrail-ml-service
  • Update URLs to point to agenticants.ai for consistency
  • Refine safety detection labels and remove toxicity references from queries
  • Updated black js package
  • Protect agent display names from sync overrides, improve integration services, and add reusable logo components
  • Governance UI improvements , empty state, drift alerts, profile header
  • Resolve merge artifacts and build errors
  • Add selfHostedGuardrailEnabled to getScheduleOptions response
  • Add guardrailBlockedMessage to Zod schema and upsert operations
  • Add guardrailBlockedMessage to getConfig response
  • Resolve TypeScript build errors in PiiPciTracesTable
  • Address PR audit round 4 (critical + high priority)
  • Address PR audit round 3 (security & correctness)
  • Address PR audit round 2
  • Regenerate pnpm-lock.yaml after develop merge
  • Address PR audit critical issues
  • Integrate Zod resolver in Evidence Collection Page and enhance search functionality in Drift Detection Service
  • Update Evidence Collection Page to set default attestation values to false and improve button layout in Configuration Summary
  • Update text to use HTML entities for apostrophes in governance components
  • Security, performance, and DRY fixes across risk tab and dashboard pages
  • Handle potential null riskScore in StatusBadge tooltip and update catchAndLog return type for error handling
  • Redesign integrations page with compact cards, table-based mappings, breadcrumbs, and delete confirmation
  • Streamline integration icon handling in DriftNotificationButton component
  • Enhance table layout and badge styling in compliance and validation sections
  • Update compliance status terminology and enhance agent listing functionality
  • Improve governance UI components for evidence collection and compliance display
  • Update governance server services and router for compliance reporting
  • Update AgentEvidence components for improved pagination and status handling
  • Update image tags to v0.1.53 for consistency across services and add start/stop commands for ethics-bias-detector
  • Correct per-framework coverage deduplication and cleanup dead code
  • Update risk summary card label and fix sign-up and assign pages
  • Resolve React hooks ordering violation in PiiPciTracesTable
  • Improve framework inventory table and stat card styling
  • Polish governance UI components and simplify stat card
  • Correct per-framework coverage calculation and governance service improvements
  • Build error
  • Pin dd-trace and @prisma/client versions, add DuoGuard to Helm deployments
  • Replace broken plugin-based code review with official multi-agent prompt
  • Resolve web build errors after develop merge
  • Correct installation command for Python dependencies in Dockerfile
  • Fixed some ui changes of connector
  • Resolved merge conflicts
  • Set ethics detector maxTextLength to 8192 for g4dn GPU compatibility
  • Fixed servicenow integration
  • PagerDuty UI improvements and incident agent resolution
  • Add ants-ethics-bias-detector-llama-1b to ECR cleanup repositories
  • Resolve TypeScript enum type errors in ai-governance module
  • Use gettext instead of gettext-base for Alpine Linux
  • Add gettext-base to apk install for runtime dependencies
  • Add SSL/TLS support for RabbitMQ connections
  • Fix RabbitMQ SSL connection and add HuggingFace token
  • Improve performance, type safety, and code quality
  • Standardize ClickHouse query parameterization
  • Added violation count feature
  • FinOps: Exclude failed models from cost/latency optimization recommendations
  • Remove duplicate pendingJobUpdates.push() in eval-batch-processor
  • Fixed the configmap bug
  • Completely disable system logs using remove="1" attribute
  • Remove ENGINE=Null config incompatible with ClickHouse 25.10
  • Disable ClickHouse system logs and increase resource limits
  • Add missing variables in select-models-form from merge conflict
  • Add missing uniqueId and pricing variables in select-models-form
  • Remove duplicate configCounts variable declaration in finops router
  • Fixed duplicate code
  • Update tsconfig to exclude additional script directories
  • Resolve build errors and remove deprecated code
  • Handle eval batch failures by marking PENDING job executions as ERROR
  • Add MapPin to codespell ignore list
  • Removed unused code and improved error handling
  • AI Command Center: Adjust layout and improve theme switcher functionality in AgentDiscoveryBar and ThemeSwitcher components
  • Skip storing failed traces in ClickHouse and adjust metadata handling
  • Add failed trace count tracking and model token limit handling in RabbitMQ and OpenAI clients
  • Implement public wrapper for publishing FinOps completion messages in batch checkers
  • Improve error logging by safely extracting message properties in FinOps job execution
  • Address PR review issues - security, race conditions, memory leaks, and typing
  • Use ClickHouse traces for agent discovery in Slack/Jira
  • Pricing lookup bug in cost-calculator and RabbitMQ TLS hostname verification
  • Resolve ClickHouse disk space issue and increase volume to 100Gi
  • Resolve .gitignore conflict and format code
  • Improve recommendation-engine reliability and batch inference support
  • Improve worker and trace-agent reliability and cleanup dead code
  • FinOps: Improve status tracking reliability with upserts and retries
  • AI Command Center: Optimize queries and graceful error handling
  • Simplify recommendation-engine Dockerfile to properly copy Prisma client
  • Clean model name display across FinOps pages
  • AI Command Center: Replace hardcoded API URL with environment variable for better configuration
  • AI Command Center: Standardize integration filter name to match API expectations
  • AI Command Center: Resolve critical filter operator and type mismatch issues
  • Add missing Bash permissions for git and gh commands in settings.json
  • Enhance code review checklist with detailed criteria for correctness, architecture, security, performance, testing, readability, documentation, and operations
  • Load environment variables from root .env
  • AI Command Center: Improve price formatting UX
  • Use pipe delimiter to avoid conflicts with model names containing colons
  • Include connectionId in comparison models URL params for FinOps optimizer
  • FinOps: Handle model names with colons in LLM optimizer
  • AI Command Center: Add caching and validation for scale
  • AI Command Center: Fix filtering bugs, SSR issues, and skeleton UI
  • Add filterRing to codespell ignore list
  • FinOps: Correct TypeScript type errors in FinOps components
  • FinOps: Use only best model per recommendation for savings calculations
  • Enable RabbitMQ consumers for LLM Judge evaluations
  • FinOps trace consistency - ensure identical trace selection across baseline and comparison evaluations
  • Platform: Fix agent card empty models background and filter button height
  • FinOps: Use upsert for model comparison to handle race conditions
  • Azure OpenAI batch inference - fix modelName handling and cost calculation
  • Ui changes-AI optimizer screens & added locking mechanism for azure cron consumer
  • Formatting issues
  • Add cost calculation to batch clients
  • Fixed helm chart
  • Changed-google-models-selection-logic
  • Update image tags to v0.1.12 for production deployment
  • Enable refreshable MVs via connection parameter instead of SET command
  • Add IST (Indian Standard Time) to codespell ignore list
  • Add *.pem to codespell skip list
  • Skip .pem files in codespell to avoid false positives
  • Exclude .pem files from codespell checking
  • Remove accidentally added golang-migrate files and format code
  • Pass connection ID with comparison models for correct provider selection
  • FinOps: Use scientific notation for small cost values
  • FinOps: Pass traceCount from sampling slider to worker batch action
  • Ensure agentName is always stored in lowercase snake_case for consistency
  • FinOps: Remove eval completion queue consumer from worker
  • Complete PostgreSQL SSL certificate verification implementation
  • Correct showErrorToast function calls to use positional parameters
  • Add type guards for useBatchInference property access
  • Skip latency optmizer when batch inference is enable
  • Update bedrock batch inference code
  • FinOps: Exclude traces with error-level observations
  • Remove NODE_TLS_REJECT_UNAUTHORIZED workaround in Jira integration
  • Handle non-array filter data in evaluator table
  • Add enabled flag to trace-agent-cron helm chart
  • ECR cleanup missing region parameter and tag splitting
  • Fixed web build issue
  • Remove trailing dash in image tags for tagged commits
  • Add missing @aws-sdk/client-iam dependency to web package
  • Finops job configuration id not passing properly - issue is fixed
  • Fixed incorrect message format issue for finops-job-execution queue
  • Added filter for the baseline model
  • Enhance trace reuse by fetching original baseline traces for input data
  • Fixed json formating for the ai recommendation
  • Enhance batch processing by integrating ClickHouse for original trace retrieval and adding model filtering for FinOps comparisons
  • Checkbox is updated to decide between real-time and batch-api implementation
  • Added batch-api for the finops flow also
  • Resolve TypeScript build errors for batch inference feature
  • Add missing dependencies and regenerate Prisma client
  • Standardize batch inference config paths in Helm charts
  • Improve batch inference validation logic
  • Update null checks for cost and latency recommendations
  • Fixed recommendation UI issue
  • Enable trace-agent deployment and fix KEDA ScaledObject
  • Rewrite agent discovery to query ClickHouse dynamically
  • Add defensive array check for evaluatorIds in execution history
  • Add defensive array checks for recommendation models in finops router
  • Add defensive null check for JobExecution array in finops router
  • Add defensive null checks for evaluators array in job details page
  • Add defensive null checks for agent mappings to prevent .map() crash
  • Add JSX support to trace-agent tsconfig and update lockfile
  • Update UI components and improve finops, and RBAC features
  • Prevent duplicate validation toasts and improve loading state handling in evaluator selection
  • Handle missing job configurations gracefully in finops comparison completion
  • Improve Prisma type imports and enhance trace-agent cost calculation
  • Resolve trace-agent Docker build failures
  • Quote inline cache parameters to prevent build errors
  • Replace AWS CLI verification with simple success message
  • Correct spelling errors flagged by codespell
  • Update codespell configuration to ignore generated files and valid acronyms
  • Exclude error_handling traces from FinOps cost calculations
  • Temporarily bypass data verification in ClickHouse due to client issues
  • Explicitly specify database in ClickHouse client config and queries
  • Improve RabbitMQ connection verification and ensure correct database in ClickHouse config
  • Remove non-existent total_cost column from verification query
  • Enable SSL/TLS for PostgreSQL connections in Cost Calculator
  • Correct RDS endpoint and handle external database hostnames in ConfigMap
  • Enhance RabbitMQ connection handling with connection string support
  • Parse RabbitMQ URL in trace-worker config
  • Update pnpm-lock.yaml to match package.json next-auth version
  • Bug fixes
  • Correct Dockerfile paths for trace-orchestrator and trace-worker builds
  • Add documentation links to quickstart examples
  • Resolve TypeScript errors in recommendation scripts
  • Jira integration is completed
  • Added jira board integration
  • Update trace-orchestrator Dockerfile paths in Makefile
  • Implement comprehensive credit authorization for all ingestion endpoints
  • Fixed cost calculation issue
  • Resolve React version conflicts, onboarding survey bug, and remove hardcoded values
  • Resolve build issues and update dependencies
  • Resolved build issue
  • Fixed trace orchestrator issues
  • Resolved pnpm issue
  • Added region to queue
  • Formating the document
  • Updated k8s scripts
  • Fixed build issue of worker
  • Remove owner field from dashboard upsert script
  • Correct Docker build context in docker-compose.agentic-ants.yml
  • Correct base image reference in Dockerfiles
  • Remove .turbo directory copy from web Dockerfile

Security

  • Security hardening

July 2026

8 releases.

v0.1.90

Released 2026-07-05.

Fixes

  • Activate GHA BuildKit cache + drop pure-ordering job dependency

v0.1.89

Released 2026-07-05.

Improvements

  • Parallelize PR validation image builds (same matrix as deploy flow)

v0.1.88

Released 2026-07-05.

Maintenance release: dependency updates and internal improvements.

v0.1.87

Released 2026-07-05.

Improvements

  • Split image builds into parallel jobs + enable BuildKit layer cache

Fixes

  • Prevent push-chart from masking helm push failures

v0.1.86

Released 2026-07-05.

Maintenance release: dependency updates and internal improvements.

v0.1.85

Released 2026-07-05.

Maintenance release: dependency updates and internal improvements.

v0.1.84

Released 2026-07-05.

Maintenance release: dependency updates and internal improvements.

v0.1.83

Released 2026-07-05.

Maintenance release: dependency updates and internal improvements.

May 2026

1 release.

v0.1.75

Released 2026-05-31.

Features

  • Shadow AI: Add Uploads/Tools and Monitored By columns to user Recent Activity
  • Shadow AI: Implement drag-and-drop sorting for navigation items and groups
  • Shadow AI: Versioned release filenames + unified device capability tiles
  • Shadow AI: Update device node height and enhance provider lookup functionality
  • Shadow AI: Add project folder support to metrics and UI components
  • Shadow AI: Enhance data handling and reporting
  • Shadow AI: Add 'kind' property to Windows tool configurations
  • Shadow AI: Tabbed apps UI, parser session table, Windows agent settings & identity fix
  • Shadow AI: Add Microsoft 365 Copilot + promote consumer Copilot to intercept
  • Telemetry: CrewAI Enterprise + generic gen_ai semconv parser
  • Add Windows detection for ChatGPT, Microsoft Copilot, GitHub Copilot CLI, Notion, Grammarly
  • Add Aider CLI Windows detection rules
  • Shadow AI: Unify desktop + browser deployment UI behind segmented toggles
  • Shadow AI: Enrich CLI sessions with parser data and file upload tracking
  • Shadow AI: Enhance CLI support and improve uninstall script for Windows agent
  • Shadow AI: Identity v2 cutover + UX cleanup across registry, users, devices
  • Shadow AI: Windows agent ingestion, device dedup, registry fixes
  • Shadow AI: Accept windows-agent as a valid client kind
  • Shadow AI: Internal Apps in extension config + piggyback refresh
  • Shadow AI: Icon + Powered by X plumbing across UI
  • Shadow AI: Registry rebrand , Internal Apps + My Nominations tabs
  • Shadow AI: Internal Apps CRUD + nomination dedup + icon upload tRPC
  • Shadow AI: Icon storage + upload pipeline
  • Shadow AI: Layer 2 Prisma guard + RLS GUC context
  • Shadow AI: Model breakdown + activity graph views

Fixes

  • Shadow AI: Add missing uploads field to app-detail recentSessions type
  • Shadow AI: Update Mistral domain to correct URL
  • Shadow AI: Tab persistence, detail uploads, detection method, agent liveness
  • Shadow AI: Consistent interaction/session counts across all dashboard surfaces
  • Shadow AI: Complete desktop AX promptDetection spec for macOS agent
  • Shadow AI: Drop Bing from Copilot registry; count token-only CLI sessions
  • Shadow AI: Dedup recent activity rows, strip Parser prefix, fix empty context fallback
  • Shadow AI: Simplify user detail, fix nav collapse, fix settings routing
  • Shadow AI: Parser-only CLI session counts + correct tool/token display
  • Shadow AI: Unify interaction counting across all dashboard views
  • Deployment: Purge stale subchart .tgz archives + preflight to prevent recurrence
  • Shadow AI: Prevent close icon overlapping Open in new tab button in file events sheet
  • Shadow AI: Resolve missing logos for GitHub Copilot CLI and Microsoft 365 Copilot
  • Telemetry: Tag CrewAI agents with framework=crewai so the UI shows the logo
  • Repair macOS aider classifier + split gemini_cli from gemini web row
  • Shadow AI: Migrate Edge extension to Microsoft Edge Add-ons store
  • Shadow AI: Resolve code review issues on CLI parser enrichment
  • Remove windows binaryTarget to unblock Docker builds
  • Reduce web build Node.js heap from 8192 to 6144 MB to prevent OOM
  • Shadow AI: Replace broken Mode tab toggle in WindowsSection with PlatformMdmDisclosure collapsible
  • Shadow AI: Code review fixes , registry admins, icon upload procedure, custom app extended fields
  • Shadow AI: Hotfix-6 , Chrome extension migrated from S3 to Chrome Web Store
  • Shadow AI: Remove redundant overflow styles in ShadowAiUserDetail component
  • Shadow AI: Serve enroll/unenroll scripts as text/plain to bypass Chrome download block
  • Shadow AI: Flush cfprefsd in enroll + unenroll scripts
  • Shadow AI: V0.1.67 hotfix-1 , daily-chart TZ, unenroll completeness, guardrail probe
  • Shadow AI: Windows agent support across UI + data correctness icon + provider display names
  • Shadow AI: Split appHost/apiHost in enrollment allowlist + installer
  • Shadow AI: Correct sidebar active state and browser title on nested pages

April 2026

7 releases.

v0.1.64

Released 2026-04-27.

Features

  • Shadow AI: Enhance enrollment script with enroll/unenroll modes and improve client count logic
  • Adopt langfuse pricing tiers + refresh model catalog
  • AI Command Center: Add Inactive status filter and demote inactive agents
  • Shadow AI: Deployment & Keys rework , shared enroll-script Step 0 + inline MDM
  • Shadow AI: Add post-enrollment redirect to Shadow AI dashboard
  • Shadow AI: Per-session active time from hostFlow + session column
  • Shadow AI: Use vendorWebsite as favicon fallback when apiHosts is absent
  • Shadow AI: Include CLI session data across Overview + User Detail aggregations
  • Rebrand plans to Agentic Observability & Guardian AI
  • Shadow AI: Gate pages behind enterprise + refine focus-time calc
  • Shadow AI: Add Apple logo icon component and update macOS section icon
  • Shadow AI: Unify browser/macOS sessions across app & user views
  • Shadow AI: Success panel on enroll page + hostname normalization
  • Shadow AI: Add SourceBadge component and update app detail source tracking
  • Shadow AI: Browser/macOS read-layer unification + registry UX cleanup
  • Enhance risk level handling in shadow AI router
  • Changed signin flow for extension and agent
  • Implement collapsible sections for navigation groups and update route visibility fix: remove unused cost projections from AgentDetailPanel and PerformanceTab components style: add collapsible animations to Tailwind configuration
  • Shadow AI: Add settings overview, deployment, and policy components
  • Shadow AI: Add per-user CLI and browser activity endpoints; enrich provider data with sanctions and categories
  • Shadow AI: Implement CLI tool detail sheet and integrate with CLI tools section
  • Remove legacy Shadow AI activity and session models, and related API endpoints
  • Shadow AI: Add device/file detail sheets, cli tools section and wire new components into pages
  • Shadow AI: Add shared badges, section primitives and risk/confidence constants
  • Shadow AI: Add clients and registry public APIs, update events ingestion
  • Shadow AI: Add clickhouse read/write services, banned key guard and extend router
  • Shadow AI: Implement custom app management with creation, editing, and listing functionalities
  • Shadow AI: Enhance Shadow AI components with sanction management and app detail features
  • Shadow AI: Implement Shadow AI app detail sheet and enhance related components with sanction management
  • Shadow AI: Add app sanction posture (constants, badge, card)
  • Shadow AI: Enhance layout and navigation for Shadow AI features, including new routes and improved component integration
  • Shadow AI: Add apps, devices, reports, users pages integrate into layout, routes, and command menu
  • Shadow AI: Update feature tables, overview, reports, user detail
  • Shadow AI: Add Shadow AI components and shared utilities
  • Add Shadow AI favicon proxy with S3 caching
  • Implement utility functions for Shadow AI including collapseTopN, formatting, debouncing, and user display metadatafeat: implement utility functions for Shadow AI including collapseTopN, formatting, debouncing, and user display metadata
  • Add constants for shadow AI features including app providers, categories, device health, chart styles, and time ranges
  • Add type definitions for shadow AI features including activity, app, device, session, user, and overview
  • Added apis for shadow ai
  • Enhance privacy notice links and Update sorting options to prioritize 'last_active'
  • Implement integration action components for Jira, Slack, and Linear in FinOpsRecommendationsWithSavings
  • Remove VERSION display handling and update font constants to include Poppins
  • Add new integration documentation links for CrewAI, Slack, Jira, and Linear
  • Update documentation links in AnnotationQueuesOnboarding and DatasetsOnboarding components
  • Add isTableAction prop to DeleteEvalConfigButton in EvaluatorTable and FinOpsJobConfigurationDetails
  • Integrate Linear issue creation functionality in FinOps recommendations
  • Improve bot icon implement agent tags context and update BotIcon to use agentId for platform resolution.
  • Add SIL Open Font License text for font software compliance
  • Add font switching feature and optimize image caching
  • Add VERSION file support and improve layout spacing in components

Fixes

  • Pre-PR review findings (gzip bomb, ESO optional fields, tier upsert)
  • Use kebab-case subchart names for helm --set image.tag
  • Create default PricingTier when upserting flat prices
  • AI Command Center: Distinguish sub-tenth-of-a-cent costs from zero in formatCurrency
  • Shadow AI: Align Session Timeline device grouping with Devices page
  • Shadow AI: Debounce favicon loads + clip user-detail x-overflow
  • Shadow AI: Merge per-person across installs, fix file uplaod issue, show checkin history for browser devices
  • Shadow AI: Scope user-detail Files drill-down to the clicked user
  • Shadow AI: Correct MDM plist key prefix + add dashboard indexes
  • Adjust column size and styling for Drift Alerts Table and Notification Button
  • Updated ai registry settings
  • Updated schema for the browser tracking
  • Linear integration missing schema module and enhance guardrail policies
  • Build errors.
  • Open external agenticants.ai links in new tab
  • Make pagerduty_incidents.integration_id nullable to support onDelete: SetNull

Security

  • Security hardening

v0.1.63

Released 2026-04-04.

Maintenance release: dependency updates and internal improvements.

v0.1.62

Released 2026-04-03.

Fixes

  • Update spacy from yanked 3.8.4 to 3.8.14 in guardrail-ml-service

v0.1.61

Released 2026-04-03.

Fixes

  • Update spacy from yanked 3.8.4 to 3.8.14 in guardrail-ml-service

v0.1.60

Released 2026-04-03.

Fixes

  • Update URLs to point to agenticants.ai for consistency

v0.1.59

Released 2026-04-03.

Maintenance release: dependency updates and internal improvements.

v0.1.58

Released 2026-04-02.

Features

  • Add Linear integration to AI Command Center agent detail
  • Add Linear notification support for compliance, detection, and governance
  • Add Linear support to agent project mappings with 3-way merge
  • Add Linear settings page, connection card, and logo assets
  • Add Linear tRPC router, OAuth handler, and API route
  • Implementation of guadrail feature
  • Enhance governance compliance tracking and UI components with evidence coverage and attestation status
  • Enhance Framework Compliance Summary and Dashboard with Agent Compliance Metrics
  • Enhance risk dashboard and governance components
  • Enhance GovernanceProfileService to track requirement completion and agent compliance details
  • Enhance FrameworkComplianceCard with donut chart and tooltips for compliance status
  • Update risk dimension descriptions and tooltips for clarity and accuracy
  • Implement compliance record management in GovernanceEvidenceService and sync with GovernanceProfileService
  • Add expand/collapse functionality to ConfigurationSummary and PolicyConfigForm components
  • Enhance RiskDashboardPage and RiskSummaryCards with improved navigation and link functionality
  • Add drift alert incident handling for PagerDuty and ServiceNow
  • Add CellTooltip component and integrate it into DriftAlertsTable and DriftNotificationButton for improved tooltip functionality
  • Enhance UI of governance components with improved styling and action slots
  • Add BotIcon component to various governance pages and enhance UI elements
  • Enhance drift alerts functionality and UI across governance sections
  • Enhance drift alerts functionality with pagination, filtering, and improved skeleton UI
  • Add Drift Alerts components and utilities for governance monitoring

Fixes

  • Refine safety detection labels and remove toxicity references from queries
  • Updated black js package
  • Protect agent display names from sync overrides, improve integration services, and add reusable logo components
  • Governance UI improvements , empty state, drift alerts, profile header
  • Resolve merge artifacts and build errors
  • Add selfHostedGuardrailEnabled to getScheduleOptions response
  • Add guardrailBlockedMessage to Zod schema and upsert operations
  • Add guardrailBlockedMessage to getConfig response
  • Resolve TypeScript build errors in PiiPciTracesTable
  • Address PR audit round 4 (critical + high priority)
  • Address PR audit round 3 (security & correctness)
  • Address PR audit round 2
  • Regenerate pnpm-lock.yaml after develop merge
  • Address PR audit critical issues
  • Integrate Zod resolver in Evidence Collection Page and enhance search functionality in Drift Detection Service
  • Update Evidence Collection Page to set default attestation values to false and improve button layout in Configuration Summary
  • Update text to use HTML entities for apostrophes in governance components
  • Security, performance, and DRY fixes across risk tab and dashboard pages
  • Streamline integration icon handling in DriftNotificationButton component
  • Enhance table layout and badge styling in compliance and validation sections
  • Update compliance status terminology and enhance agent listing functionality
  • Improve governance UI components for evidence collection and compliance display
  • Update governance server services and router for compliance reporting
  • Update AgentEvidence components for improved pagination and status handling

March 2026

5 releases.

v0.1.54

Released 2026-03-26.

Features

  • Add Cost Performance Metrics and Latency & Analytics pages with routing updates
  • Add guardrail policy configuration to agent detail and overview tabs
  • Add ServiceNow/PagerDuty integration badges, fix risk score consistency, and UI improvements
  • Add guardrail violations and drift alerts filters, update status to risk-score based
  • Implement composite risk score and unified risk indicators for AI Command Center

Fixes

  • Handle potential null riskScore in StatusBadge tooltip and update catchAndLog return type for error handling
  • Redesign integrations page with compact cards, table-based mappings, breadcrumbs, and delete confirmation
  • Update image tags to v0.1.53 for consistency across services and add start/stop commands for ethics-bias-detector

v0.1.53

Released 2026-03-22.

Features

  • Fix build errors update time range options for risk dashboard and drift alerts
  • Enhance compliance violation metrics and drift alert queries with agent filtering
  • Enhance risk and compliance dashboard with agent name handling, improved navigation, and performance optimizations
  • Enhance risk dashboard components with project ID handling, improved drift alert labeling, and performance optimizations in data queries
  • Enhance risk compliance dashboard with project ID handling, error management, and compliance violation schema
  • Add Risk Dashboard components and hooks
  • Implement risk dashboard feature with tRPC router, computation service, and validation schemas
  • Implement risk compliance dashboard with heatmap and drift alerts overview

Fixes

  • Correct per-framework coverage deduplication and cleanup dead code
  • Update risk summary card label and fix sign-up and assign pages
  • Resolve React hooks ordering violation in PiiPciTracesTable
  • Improve framework inventory table and stat card styling
  • Polish governance UI components and simplify stat card
  • Correct per-framework coverage calculation and governance service improvements
  • Build error
  • Pin dd-trace and @prisma/client versions, add DuoGuard to Helm deployments

v0.1.52

Released 2026-03-13.

Maintenance release: dependency updates and internal improvements.

v0.1.51

Released 2026-03-10.

Features

  • Improve RBAC UI members page in org and project settings panel
  • Enhance batch processing and ClickHouse integration
  • Implement batch message schema and publishing for DuoGuard detection
  • Add driftAlert to AuditableResource type
  • Strengthen governance and drift management with runtime validation, automated policy activation, and integrated alerting
  • Improve AI Policy UI consistency and loading states

Fixes

  • Replace broken plugin-based code review with official multi-agent prompt
  • Resolve web build errors after develop merge
  • Correct installation command for Python dependencies in Dockerfile
  • Fixed some ui changes of connector

v0.1.50

Released 2026-03-05.

Features

  • Add LLM Inventory, My Requests, and Violations previews
  • Enhance ethics bias entity handling and validation in PiiPciTracesTable
  • Production-ready connectors with graceful shutdown, helm improvements, and v0.1.48
  • Add scripts and configuration for managing enterprise plan entitlements
  • Wrap AI compliance and command center components with EnterpriseFeatureGate for plan restrictions
  • Implement enterprise features and entitlement checks
  • Update favicon and icon assets for improved branding
  • Update SignIn component to support dark mode for AgenticAnts logo
  • Update AgenticAnts logo with light and dark mode support
  • Low code now code connector
  • Improve compliance section ui and enhance chart color semantics for better data visualization
  • Improve ethics/bias detection accuracy with model-based bias detection
  • Added llmn trace connector for bedrock and vertex

Fixes

  • Resolved merge conflicts
  • Set ethics detector maxTextLength to 8192 for g4dn GPU compatibility

February 2026

3 releases.

v0.1.43

Released 2026-02-21.

Features

  • Added credit consumption for ai policies
  • Added credit consumption logic for ai policies
  • AI Governance: Enhance framework compliance data retrieval and UI with filtering options
  • AI Governance: Add Export & Audit endpoints for compliance reports and audit logs - Introduced listReports and getAuditLog endpoints in the AI Governance router. - Implemented corresponding methods in the GovernanceProfileService to handle report listing and audit log retrieval with filtering and pagination. - Added validation schemas for the new endpoints: ListReportsInputSchema and GetAuditLogInputSchema. - Updated the governance router documentation to reflect the new endpoints and their functionalities.
  • AI Governance: Implement dashboard statistics and recent activity endpoints with API integration
  • AI Governance: Add notification fields and feedback mechanisms for Slack and Jira integration
  • AI Governance: Enhance agent mapping integration for Slack and Jira in governance components
  • AI Governance: Enhance governance schemas and database structure with new evidence form fields and agent listing functionality
  • AI Governance: Pass agentName to RuntimeValidationSection for trace navigation
  • AI Governance: Enhance framework compliance mapping and add latest attestation logic
  • AI Command Center: Multi-select compliance notification modal, icon/tooltip fixes, and ServiceNow/PagerDuty removal
  • AI Governance: Update compliance framework identifiers to uppercase
  • AI Governance: Add schema + governance utilities

Fixes

  • Fixed servicenow integration
  • PagerDuty UI improvements and incident agent resolution

v0.1.40

Released 2026-02-05.

Features

  • Add PagerDuty incidents page with pagination and agent/status filters
  • Add PagerDuty integration for compliance alerting

Fixes

  • Add ants-ethics-bias-detector-llama-1b to ECR cleanup repositories

v0.1.39

Released 2026-02-05.

Features

  • Add Role-Based Access Control (RBAC) documentation for AI Governance feature
  • Add Role-Based Access Control (RBAC) documentation and update API access details in AI Compliance docs
  • Update request status handling and UI components in AvailableRequestsPage and MyRequestsPage
  • Update MyRequestsPage to clarify status labeling and format date display
  • Enhance ExportAuditPage UI with improved tab styling
  • Add edit functionality and pagination to Drift Alerts section
  • Remove unused activity filters and related components from GovernanceDashboardPage
  • Added servicenow integration
  • Add filter removal functionality to PopoverFilterBuilder and InventoryFilters
  • Remove VIEWER role from AI Compliance API specification
  • Add script to import new models from JSON and update Excel inventory
  • Enhance Evidence Collection and AI Compliance tabs based on user roles
  • Add ethics_bias_labels column to pii_pci_config table
  • Add Request Changes and Request Creation modals for AI Governance
  • Add comprehensive API specification doc for AI Compliance feature
  • Implement assignment route for governance profiles and update navigation paths
  • Implement recent activity tracking with filters and detailed activity types
  • Enhance framework compliance section with additional requirements and framework badges
  • Adjust column widths and add Request column in Agent Evidence Table

Fixes

  • Resolve TypeScript enum type errors in ai-governance module
  • Resolved merge conflicts

January 2026

10 releases.

v0.1.38

Released 2026-01-30.

Maintenance release: dependency updates and internal improvements.

v0.1.36

Released 2026-01-30.

Features

  • Configure ethics-bias-detector with GPU acceleration
  • Add Slack and Jira integration buttons to Agent Evidence Table
  • Add components for governance tracking and compliance management - Implemented CoverageProgressBar for visualizing evidence coverage. - Created DriftAlertBadge to display drift alert counts with severity styling. - Developed FrameworkComplianceCard to show compliance status for frameworks. - Added GovernanceEmptyState for displaying when no governance data is available. - Introduced GovernanceLoadingState for loading skeletons in governance components. - Created GovernanceStatCard for reusable stat cards in the Agent Governance dashboard. - Added RiskLevelBadge to display risk classification levels. - Re-exported shared components for easier imports in the governance module. - Defined constants for evidence categories, compliance frameworks, risk levels, and drift types. - Established types for evidence items, governance profiles, drift alerts, and dashboard statistics. - Implemented navigation utilities for AI compliance tabs. - Created pages for managing agent evidence, compliance reports, and audit exports.
  • Add ethics & bias detection with Llama Guard 3 1B
  • Add PostgreSQL readonly user setup script for local development
  • Add ClickHouse dictionary setup guide and PostgreSQL readonly user script
  • Enhance compliance monitoring with improved security and validation
  • Enhance agent list API with compliance data and notifications
  • Telemetry: Add organizationId to trace metadata for compliance monitoring
  • Improve UI consistency and add tooltips for LLM and framework inventory
  • Telemetry: Add organizationId to ingestion events and processing logic
  • Simplify approval UX and unify inventory UI
  • End-to-end compliance support across backend and AI Command Center
  • Enhance ApprovalCell with toggle switch and improve MultiSelectDropdownFilter with search functionality

Fixes

  • Use gettext instead of gettext-base for Alpine Linux
  • Add gettext-base to apk install for runtime dependencies
  • Add SSL/TLS support for RabbitMQ connections
  • Fix RabbitMQ SSL connection and add HuggingFace token
  • Improve performance, type safety, and code quality
  • Standardize ClickHouse query parameterization
  • Added violation count feature

Security

  • Security hardening

v0.1.33

Released 2026-01-13.

Features

  • Add phi detection
  • PHI implementation

Fixes

  • FinOps: Exclude failed models from cost/latency optimization recommendations
  • Remove duplicate pendingJobUpdates.push() in eval-batch-processor

v0.1.32

Released 2026-01-11.

Features

  • Implement model processing status updates for FinOps evaluations and trace generation
  • Enabled gliner spacy
  • Enhance UX/UI with accessibility and visual improvements
  • Add Framework Inventory UI with dual-tab interface
  • Add Framework Inventory backend and database schema
  • Enhance provider requirements section with Azure deployment details
  • Remove FinOps recommendation toggle and adjust related form fields
  • Add description field to EvalTemplate and update related components
  • Enhance model selection with pricing data and sorting functionality
  • Add LLM Inventory page and filters, enhance routing and table types, Small supporting changes
  • Add frontend UI components and hooks
  • Publish FinOps eval completion to recommendation-engine for failed models
  • Add failed trace count tracking to batch processing and completion messages
  • Enhance batch job status handling with success and error counts for Azure and OpenAI services
  • Implement backend API and business logic
  • Enhance delete confirmation dialog with structured layout and improved accessibility
  • Implement FinOps completion check and publish mechanism in evalQueue
  • Implement deduplication and model normalization for batch processing in RabbitMQ and enhance FinOps completion checks
  • FinOps: Add incremental progress tracking for LLM optimizer jobs
  • FinOps: Add UI components for granular job status tracking
  • FinOps: Implement status updates in trace-agent and worker
  • FinOps: Add tRPC endpoints for granular job status tracking
  • FinOps: Add granular status tracking for LLM optimizer jobs

Fixes

  • Fixed the configmap bug
  • Completely disable system logs using remove="1" attribute
  • Remove ENGINE=Null config incompatible with ClickHouse 25.10
  • Disable ClickHouse system logs and increase resource limits
  • Add missing variables in select-models-form from merge conflict
  • Add missing uniqueId and pricing variables in select-models-form
  • Remove duplicate configCounts variable declaration in finops router
  • Fixed duplicate code
  • Update tsconfig to exclude additional script directories
  • Resolve build errors and remove deprecated code
  • Handle eval batch failures by marking PENDING job executions as ERROR
  • Add MapPin to codespell ignore list
  • Removed unused code and improved error handling
  • AI Command Center: Adjust layout and improve theme switcher functionality in AgentDiscoveryBar and ThemeSwitcher components
  • Skip storing failed traces in ClickHouse and adjust metadata handling
  • Add failed trace count tracking and model token limit handling in RabbitMQ and OpenAI clients
  • Implement public wrapper for publishing FinOps completion messages in batch checkers
  • Improve error logging by safely extracting message properties in FinOps job execution
  • Address PR review issues - security, race conditions, memory leaks, and typing
  • Resolve .gitignore conflict and format code
  • Improve recommendation-engine reliability and batch inference support
  • Improve worker and trace-agent reliability and cleanup dead code
  • FinOps: Improve status tracking reliability with upserts and retries

Security

  • Security hardening

v0.1.29

Released 2026-01-05.

Fixes

  • Use ClickHouse traces for agent discovery in Slack/Jira
  • Pricing lookup bug in cost-calculator and RabbitMQ TLS hostname verification
  • Resolve ClickHouse disk space issue and increase volume to 100Gi

v0.1.27

Released 2026-01-04.

Features

  • Add /pr-audit command for automated pre-PR code review

Fixes

  • AI Command Center: Optimize queries and graceful error handling
  • Simplify recommendation-engine Dockerfile to properly copy Prisma client
  • AI Command Center: Replace hardcoded API URL with environment variable for better configuration
  • AI Command Center: Standardize integration filter name to match API expectations
  • AI Command Center: Resolve critical filter operator and type mismatch issues
  • Add missing Bash permissions for git and gh commands in settings.json
  • Enhance code review checklist with detailed criteria for correctness, architecture, security, performance, testing, readability, documentation, and operations
  • Load environment variables from root .env
  • AI Command Center: Improve price formatting UX
  • AI Command Center: Add caching and validation for scale
  • AI Command Center: Fix filtering bugs, SSR issues, and skeleton UI

v0.1.26

Released 2026-01-02.

Fixes

  • Clean model name display across FinOps pages

v0.1.25

Released 2026-01-02.

Features

  • Add local hybrid development setup with ephemeral bastion

Fixes

  • Use pipe delimiter to avoid conflicts with model names containing colons

v0.1.24

Released 2026-01-02.

Maintenance release: dependency updates and internal improvements.

v0.1.22

Released 2026-01-01.

Fixes

  • Include connectionId in comparison models URL params for FinOps optimizer

December 2025

20 releases.

v0.1.21

Released 2025-12-31.

Fixes

  • FinOps: Handle model names with colons in LLM optimizer

v0.1.20

Released 2025-12-30.

Features

  • AI Command Center: Implement dynamic theme system with variant-based styling
  • AI Command Center: Add themes with CSS variables
  • Upgrade ClickHouse image to version 25.10 for stability of Refreshable Materialized Views
  • FinOps: Add compact currency formatting for savings display
  • FinOps: Disable comparison models when 3 are selected

Fixes

  • Add filterRing to codespell ignore list
  • FinOps: Correct TypeScript type errors in FinOps components
  • FinOps: Use only best model per recommendation for savings calculations
  • FinOps: Use upsert for model comparison to handle race conditions

v0.1.19

Released 2025-12-28.

Maintenance release: dependency updates and internal improvements.

v0.1.18

Released 2025-12-28.

Features

  • Enhance logging infrastructure with Winston logger across services

v0.1.17

Released 2025-12-26.

Features

  • Add GIN index on tags for Agent model
  • Add script to sync agent tags from ClickHouse to PostgreSQL
  • AI Command Center: Add tag and integration filtering system
  • Add tag and integration filtering to agent APIs
  • Add tag extraction and JSONB support for agent metadata
  • AI Command Center: Enhance project metrics retrieval and filter handling with improved UI updates
  • AI Command Center: Implement dynamic cost filtering with adaptive boundaries and enhanced UI options
  • AI Command Center: Enhance UI components with improved styling, tooltips, and new features
  • AI Command Center: Enhance time range handling and metrics retrieval across agent detail, list, and project metrics APIs
  • AI Command Center: Add cost filter boundaries to project metrics for enhanced UI filtering
  • AI Command Center: Add time range support to project metrics and agent detail APIs
  • AI Command Center: Enhance agent detail and list hooks with time range support
  • AI Command Center: Add agent fields and optimize table TTLs for improved trace analytics
  • AI Command Center: Improve dashboard components
  • AI Command Center: Enhance types and constants for metrics

Fixes

  • Enable RabbitMQ consumers for LLM Judge evaluations
  • Platform: Fix agent card empty models background and filter button height
  • Formatting issues

Security

  • Security hardening

v0.1.16

Released 2025-12-25.

Fixes

  • FinOps trace consistency - ensure identical trace selection across baseline and comparison evaluations

v0.1.15

Released 2025-12-24.

Fixes

  • Azure OpenAI batch inference - fix modelName handling and cost calculation
  • Ui changes-AI optimizer screens & added locking mechanism for azure cron consumer
  • Enable refreshable MVs via connection parameter instead of SET command

v0.1.14

Released 2025-12-23.

Fixes

  • Add cost calculation to batch clients
  • Fixed helm chart

v0.1.13

Released 2025-12-21.

Fixes

  • Changed-google-models-selection-logic
  • Update image tags to v0.1.12 for production deployment

v0.1.12

Released 2025-12-20.

Maintenance release: dependency updates and internal improvements.

v0.1.11

Released 2025-12-20.

Maintenance release: dependency updates and internal improvements.

v0.1.9

Released 2025-12-20.

Maintenance release: dependency updates and internal improvements.

v0.1.10

Released 2025-12-19.

Features

  • Add scrollbar utilities for improved UI experience
  • Add API endpoints for agent detail, agent list, and project metrics in AI Command Center
  • Implement hooks for agent list, agent detail, and project metrics in AI Command Center
  • Add AgentCard, AgentCatalog, AgentDetailPanel, HeroMetrics, and AgentDiscoveryBar components
  • Add skeleton components for AgentCard, AgentDetail, AgentList, and MetricCard in AI Command Center
  • Add shared components for AI Command Center including FilterDropdown, SearchInput, SearchableFilterDropdown, SortControls, StatCard, SimpleSparkline, TimeRangeSelector, and ViewToggle
  • Add AI Command Center constants, types, and utility functions
  • Complete Azure OpenAI Batch API implementation for FinOps and LLM-as-Judge
  • Implement Azure OpenAI real-time API support for FinOps and LLM-as-Judge flows
  • Add hyphenated trace attributes to OtelIngestionProcessor
  • FinOps: Add projected savings with configurable assumption
  • Implement agent upsert logic for trace and observation events in IngestionService
  • Synchronize agent display name across agents and agent_project_mappings tables
  • Implement agents table for SDK-created agents and update related queries
  • Enhance agent model with mutable display name and update ingestion processor logic
  • Add SSL certificate mounts to web, worker, and recommendation-engine
  • Implement PostgreSQL SSL certificate verification for trace-agent and trace-cron
  • Improve formatting and logging for agent_id generation and validation in ingestion services
  • Add agent_id and agent_name fields for trace and observation processingand update the ingestion code with new implementation
  • Enhance logging for agentId and agentName extraction in OtelIngestionProcessor and IngestionService
  • Add agent_id and agent_name fields for enhanced tracking in observations
  • FinOps: Pass trace count to FinOps job based on sampling
  • Display sampled trace count on sampling slider

Fixes

  • Add IST (Indian Standard Time) to codespell ignore list
  • Add *.pem to codespell skip list
  • Skip .pem files in codespell to avoid false positives
  • Exclude .pem files from codespell checking
  • Remove accidentally added golang-migrate files and format code
  • Pass connection ID with comparison models for correct provider selection
  • FinOps: Use scientific notation for small cost values
  • FinOps: Pass traceCount from sampling slider to worker batch action
  • Ensure agentName is always stored in lowercase snake_case for consistency
  • FinOps: Remove eval completion queue consumer from worker
  • Complete PostgreSQL SSL certificate verification implementation
  • FinOps: Exclude traces with error-level observations

v0.1.8

Released 2025-12-07.

Features

  • Batch inference for llm as a judge

Fixes

  • Correct showErrorToast function calls to use positional parameters
  • Add type guards for useBatchInference property access
  • Skip latency optmizer when batch inference is enable
  • Update bedrock batch inference code
  • Fixed web build issue

v0.1.7

Released 2025-12-05.

Fixes

  • Remove NODE_TLS_REJECT_UNAUTHORIZED workaround in Jira integration

v0.1.6

Released 2025-12-03.

Fixes

  • Handle non-array filter data in evaluator table
  • Add enabled flag to trace-agent-cron helm chart

v0.1.5

Released 2025-12-02.

Fixes

  • ECR cleanup missing region parameter and tag splitting

v0.1.4

Released 2025-12-02.

Maintenance release: dependency updates and internal improvements.

v0.1.3

Released 2025-12-02.

Fixes

  • Remove trailing dash in image tags for tagged commits

v0.1.2

Released 2025-12-02.

Maintenance release: dependency updates and internal improvements.

Endpoint components

The browser extension and the Windows and macOS endpoint agents are released separately from the platform and update through your MDM or browser policy.

Browser extension

October 2026 (version 0.1.19)

  • Tracking of non AI apps

September 2026 (version 0.1.18)

  • Destroy the device key when registration is refused as blocked
  • Toolbar icon stuck amber on certified/device-key installs
  • Tag S3 uploads with the version metadata the download route reads
  • Device auth revoke handling and enrollment reliability

August 2026 (version 0.1.16)

  • Maintenance and internal improvements.

July 2026 (version 0.1.14)

  • Shadow AI: Prevent stuck re-enroll state in MDM mode + correct API host defaults

June 2026 (version 0.1.12)

  • Maintenance and internal improvements.

May 2026 (version 0.1.12)

  • Refuse upload without hardware deviceID; re-add managed_schema; v0.1.12
  • Remove Bing hosts from Microsoft Copilot entry
  • Dedup fileUpload across DOM+network, cap stream read, drop body log
  • Cache managed creds to local so enrollment survives reboot

April 2026 (version 0.1.7)

  • Add script-matches.js for content script match pattern computation
  • Self-heal connection on offline/online transitions
  • Fixed file upload logic
  • Identity and access: Thread managed device + user identity from chrome.storage.managed
  • Shadow AI: Session active-time + file-upload dedup in SW
  • Preserve apiHost input while user is typing
  • Ui chnages

Windows endpoint agent

October 2026 (version 0.1.18)

  • Require a local proxy token on the OTEL loopback listener
  • Collector enforcement , IT can switch collectors off by device policy

September 2026 (version 0.1.17)

  • Dev-mode uninstall fallback now removes the device cert too
  • Certified device permanently stuck on shared key under required mode
  • Stop reusing a device key the server has already disowned
  • Fleet enrollment reliability + route traces through device cert
  • Device auth revoke handling and enrollment reliability

August 2026 (version 0.1.10)

  • Address pre-merge review findings on PR#2
  • Keep DeviceInfoProvider reading identity after auth-triple deletion
  • Replace ineffective util:PermissionEx with a working registry ACL lock
  • Correct WiX util extension namespace and registry key ACL setup
  • Pin WixToolset.Util.wixext to the 5.x line in setup docs

July 2026 (version 0.1.10)

  • Add Grammarly mouse-click observer and update UIA handling for composer profiles
  • Add Grammarly mouse-click observer and generation counting logic

June 2026 (version 0.1.10)

  • Implement GrammarlyGO mouse-click observer for generative AI usage tracking
  • Enhance conversation session handling with ComposerId and durability
  • Capture doc-context, model, file uploads via UIA
  • Ship model + file uploads on AppSessionWire
  • App_session model_hint + file_uploads storage (schema v9/v10)
  • Implement Notion AI observer (UIA)
  • Enhance Aider CLI parser specifications and data model for improved prompt counting and token aggregation
  • Implement AiderParser for processing Aider chat history
  • Add initial implementation and specifications for Aider CLI parser
  • VS Code Copilot Chat parser (vscode_copilot_chat)
  • Multi-pattern command-line exclusion on IParser
  • Implement VS Code Copilot Chat parser with structural metadata capture
  • Add fast path for Service-only local deployment to reduce deploy time
  • Send per-file detectedAt so file uploads keep their real time
  • Prefer the open cli_session when resolving a parser's parent
  • Enhance process start time resolution for DCStart events to prevent double-counting
  • Read every active session, not just the polled pid's
  • Service ingest of UIA prompts + EventStore V8 prompt columns
  • Tray composer observer + content-free prompt counter
  • Update registry entry ID and clarify prompt counting logic in documentation
  • Update MS Copilot parser to ensure ModelHint is always null and clarify diagnostic log usage
  • Desktop-app parent-link tool fallback
  • Add MsCopilotParser (presence, mode, ring, version, uploads)
  • Add FindRecentOpens to IFileOpenTracker for upload detection
  • Add MS Copilot version + session-storage readers" -m "FileVersionInfo-based version reader (mscopilot.exe) and a LevelDB byte-scan reader for conversation mode + ring from Session Storage. Read-only structural metadata; never touches conversation content.
  • Add Microsoft Copilot desktop app parser specification and tasks
  • Enhance session discovery and event store with composer ID handling
  • Implement GeminiCliParser and GeminiSessionDiscovery for parsing Google Gemini CLI session files
  • Add initial implementation and task list for Gemini CLI parser
  • Improve GetLongSafe method to safely retrieve long values
  • Add CodexCliParser for OpenAI Codex CLI (codex_rollout_jsonl)
  • Enhance Codex CLI Parser with comprehensive implementation and validation
  • Document Codex CLI Parser with session tracking and token accounting
  • Add timestamped fallback for anonymous image uploads

May 2026 (version 0.1.10)

  • Implement conditional uninstall visibility based on environment and MDM enrollment
  • Security hardening
  • Add command-line exclusion filter for Electron child processes and improve workspace path resolution
  • Use real project folder (CWD) instead of composer UUID in parser_sessions
  • Final poll on process exit + skip ended Format C sessions
  • Read tokens from composerData.promptTokenBreakdown for Cursor v16 schema
  • C4/C7/C8 heartbeat flush implementation
  • Handle missing .copilot\ide directory in Format C reader
  • Add CopilotCliParser with lock-file discovery and events.jsonl parsing
  • Implement file selection handling in CursorParser and update related components
  • Enhance BatchUploader to resolve parent session IDs for parser sessions
  • Add Cursor IDE parser & file upload tracking
  • Surface real assembly version in checkin + tray instead of hardcoded 0.1.0
  • Refuse to load policy registry when identity fields are missing
  • Fix Trusted Signing DLIB extraction and metadata encoding on Windows
  • Registry version signal, tray menu cleanup, parser-only session fix
  • Update service registration to use IParser interface for Parser
  • Enhance uninstallation process with MSI support and improved cleanup
  • Use registry tool IDs verbatim; resolve parser tool ID at runtime
  • Add Uninstall menu item with full agent cleanup
  • Add dead-pid sweeper to close ghost-member sessions
  • Parser pipeline, ETW file tracker, CLI enrichment, crash log
  • Trigger recovery chain on system wake; reset hourly timer after chain
  • Guarantee ClientId before DI container builds , tray stays amber on fresh install
  • Demote unknown-stop from WARNING to TRACE
  • Gate ForegroundObserver on registry availability
  • Update status message for degraded state to reflect offline status and queueing
  • Enrich user profile with additional identity fields , fullName, UPN, employeeID, manager, identityResolvedFrom in UserBlock
  • Verify pipe caller identity to block event spoofing
  • CLI enrichment, session lock, tray buffer, agent-health tracking
  • Durable upload pipeline , batch uploader, retry queue, scrubber, recovery
  • SQLite EventStore, schema migrator, ring-buffer cap, org-key validator
  • Wire phase 2 registry+observers into worker and tray state
  • Foreground+process observers, batching upload, IPC event/state ops
  • AI tool registry fetcher, cache, classifier
  • Wix MSI, build script, smoke runner
  • Implementation of Ants Agent installer
  • Named-pipe host, agent worker, tray state machine
  • API client, enrollment exchange, URI parser, pipe protocol
  • Windows host-identity readers and DeviceInfoProvider
  • DPAPI key store, config store, rotating logger, policy registry

macOS endpoint agent

October 2026 (version 0.1.28)

  • Collector enforcement on macOS (disabledCollectors, agent 0.1.28)

September 2026 (version 0.1.27)

  • Root-owned shared dirs + stuck re-enroll menu state
  • Arm the checkin timer on the main run loop so hourly checkins fire
  • Write the team prefix literally in the keychain access group
  • Wrong Apple Team ID breaks device-key Keychain access
  • 6 mTLS device-auth bugs found testing #8 end-to-end
  • Certified device permanently stuck on shared key under required mode
  • Route coding-agent traces through the device certificate
  • Offer sign-in from the menu when the server refuses this device's keys
  • Device auth revoke handling and enrollment reliability

August 2026 (version 0.1.20)

  • Maintenance and internal improvements.

July 2026 (version 0.1.20)

  • Migrate default API host to antsplatform.ai domain
  • Initial commit , ants-macos-agent codebase
© 2026 ANTS Platform, Inc.Docs v1.0 · Last updated June 2026