Updates
Platform Updates
Every release across the ibl.ai platform — applications, infrastructure tooling, and the open source repositories behind them. You own all the code and the data, so what ships here is what you can run yourself.
LinkedInAll updates
47 updatesOrganization Settings, Rebuilt: Roles, Policies, Groups and Alerts
ibl.ai's organization settings were rebuilt around four sidebar groups, and Management now holds eight sub-tabs — Users, Groups, Roles, Policies, Teams, Alerts, Applications and Onboarding — six of them gated by their own permission.
Agents That Work on a Schedule: Tasks With Run Logs
A close look at the agent Tasks tab in ibl.ai: give an agent a named prompt, a start time and a cadence of once, daily, weekly or monthly, then check a log of every run with its status, its timing and the full output.
Incognito and PII Filtering for Every Agent
Every agent on the platform now has a Privacy tab with two sub-tabs: PII Filtering, which detects thirteen categories of personal information and allows, redacts, masks or blocks them before the model sees a message, and Incognito, which decides whether conversations with that agent are written to chat history and memory at all.
Platform Update — October 9, 2026
Eleven services and packages moved in a week, led by `ibl sandbox` — full lifecycle management for a KVM microVM code-execution platform — alongside per-agent configuration flags, a built-in CRM, product tours in the OS and LMS, incognito chat, agent shared memory, and Microsoft Graph sign-in for personal and work accounts.
Agent Network Policies and Secrets, Now in Settings
The allowlists and secrets that govern what an agent's virtual machine can reach now have screens: an organization-wide Virtual Machine tab holding the Network Policies and Secrets tables, and a Network Access section on each agent's Sandbox tab that binds them — four egress profiles, up to 100 hosts per policy, and up to 20 secrets per agent.
From One Prompt to a Live Paid App
Ask a Code Mode agent for an app that takes payments and it builds one from the default template with sign-in included, previews it on your machine, deploys it to a live web address on ibl.ai's hosting, and puts a paywall on your own Stripe account — with no Stripe keys pasted anywhere and nothing secret passing through the chat.
Platform Update — October 2, 2026
Thirteen services and packages moved in a week, led by LLM gateway unification: every model endpoint — OpenRouter, Vertex, Foundry, Bedrock and each provider's own API — is now a gateway, requests route to the highest-priority gateway that can serve the model, and a deployment with provider keys but no ibl.ai key works with no extra configuration.
Code Mode Now Runs Codex and Claude Code
Code Mode, the coding agent in Agentic OS, can now hand a turn to OpenAI's Codex or Anthropic's Claude Code on the subscription you already pay for — one row in the Code panel switches between them and ibl.ai's own models, and the same conversation continues across agents in the same workspace.
Agent Sandboxes: A Real Linux VM, Locked to Hosts You Allow
ibl.ai agents can now run code in a full Linux virtual machine that has no network by default — opened only to the hosts an organization allowlists, with API secrets the agent can use but never read, and runtime billed per second.
Agentic Vibe 2.0: Build an AI App From Your Coding Agent
Agentic Vibe 2.0 turns Claude Code, OpenAI Codex or Cursor into an ibl.ai app builder: one install of 154 skills, four questions, and a working AI app with sign-in, an agent, users and admins, memory, analytics and billing — shippable to the web, desktop and mobile.
Analytics: What Every Model Costs, Down to the Call
The rebuilt ibl.ai analytics puts nine tabs behind one date range and agent picker — usage, users, topics, replayable transcripts, memory, audit trails, and a Cost tab that breaks AI spend down by provider, model, user and individual call, with p50 and p95 latency per model.
Charge for Your AI App on Your Own Stripe Account
Apps built on ibl.ai can now sell access on the organization's own Stripe account: the admin picks free, one-time or monthly, clicks Connect with Stripe, and members pay in Stripe's embedded checkout — no commission, no webhooks, and no platform key inside the app.
Code Mode From Your Phone: Pair It to Your Desktop
Code Mode, the coding agent in Agentic OS, now runs from a phone: scan a QR code on the desktop app, and the phone drives the same agent editing files, running commands and committing in a workspace on your own computer — with every change still approved or denied from the chat.
Platform Update — September 25, 2026
Ten services and packages moved in seven days — PII and PHI filtering now covers in-chat file uploads end to end, including image pixel redaction and PDF rebuild, every detection is recorded in a new read-only privacy-flags audit endpoint, and the CLI moves nginx to nginx.org 1.30.5 with per-service ModSecurity enforcement.
Platform Update — September 18, 2026
Thirty-seven production releases in a week — the LLM model catalogue now auto-syncs behind a verification gate, credit is charged in real time with calls enforced against the balance, VM secrets let agents use credentials whose values never enter the sandbox, and the desktop app gains iOS and Android builds.
Platform Update — September 11, 2026
Sixty-two production releases in a week — agents now manage their own long-term memory through a tool-calling toolkit with expiring facts, admins and watchers can read learner chat history under RBAC, every AI turn exposes its retrieved documents and tool calls, and opt-in TLS lands for Redis, ClickHouse and MongoDB.
Platform Update — September 4, 2026
Sixty-nine production releases in a week — curated agent-level knowledge injected into every conversation, a standards-compliant OpenAI-compatible /v1 endpoint, an LLM registry refreshed to GPT-5.6, Claude Opus 5, Gemini 3.7, Grok 4.6 and DeepSeek V4, priority-ranked document retrieval, and ibl.ai Pro as the new default model.
Platform Update — August 28, 2026
Fifty-two production releases in a week — a platform-scoped LLM usage analytics API with p50–p99 latency and cost drill-down, an ibl.ai Pro / Fast model split, a voice transcription overhaul configurable per tenant without a deploy, on-device LLM providers in the agent settings UI, and a sixteen-setting control plane for per-chat-run VM sandboxes.
Deploy Your AI App: Web, Your Server, and the Stores
Three documented deployment paths shipped for apps built on ibl.ai: push a Next.js app to platform-hosted deployment using only your ibl.ai API key, containerize the same app for a server your team controls, or generate a Makefile and Fastlane config that submits the Tauri build to the App Store and Google Play.
Memory Administration and Conversation Exports
Administrators can now manage every user's memories and every agent's memories from one organization-level tab, including the switches that decide whether anything is captured at all — and every user can review and export their own conversations from their profile.
Projects and Workflows: Two Ways to Give an Agent Structure
Projects give a body of work its own files, standing instructions, and assigned agents, so every conversation started from it carries that context. Workflows lay a multi-step agent run out as a graph — branches, loops, guardrail checks, and human approval steps.
Spend Limits: Enforceable Ceilings on AI Cost at Three Scopes
LLM spend caps now exist at three scopes — the whole workspace, a single agent, and one person on one agent — each with its own reset period, an enforcement mode that blocks or alerts, and near-limit alert thresholds. Enforcement is server-side.
Platform Update — August 21, 2026
Forty production releases in a week — Code Mode gains agent skills, sandboxed execution, Stripe and Vercel deploy targets; per-agent spend caps land in the OS app; the CLI ships a 6.0.0 packaging refactor with opt-in database TLS; and the learner progress report is rewritten around an LLM-written summary.
Platform Update — August 14, 2026
Forty production releases in a week across the OS and LMS apps, the dm and auth services, the CLI, and six shared packages — Code Mode tool execution, agent memory controls, MCP server management, and a new audit log surface.
Code Mode: From One Prompt to a Running App
Code Mode is the coding agent built into the Agentic OS chat box. One prompt scaffolds a Next.js app from the Agentic Vibe starter, drops in pre-built ibl.ai components, installs dependencies, and runs it locally — with every tool call gated behind an explicit approval.
Agent Skills: Reusable Playbooks Your Agents Load On Demand
Agent Skills ship as a first-class surface: an organization-wide skill catalog, per-agent assignment, private skills, attachable file resources, and a / picker in chat that lets users invoke a skill by name.
New Testimonial: U.S. Department of Defense, Defense Industrial Base
David Heckman, PMP at the U.S. Department of Defense, on ibl.ai's work supporting an advanced manufacturing initiative for the Defense Industrial Base — now live as a highlighted testimonial on the case studies page and the home page.
User Profile and Organization Settings
Two walkthroughs of the panels behind the top-right menu in Agentic OS: the user profile, where each person controls what the AI remembers and how conversations are recorded, and organization settings, where admins brand the tenant, add their own model keys, and issue API credentials.
An AI Recruitment Agent That Fills the CRM
A Recruitment agent that searches social platforms for people publicly signaling interest in a program, qualifies each profile against the school's ideal student profiles, deduplicates against everyone it has already seen, and writes the survivors into the CRM with scores. One run read 738 Instagram posts in 45.8 seconds and recorded 13 qualified prospects for $1.70.
Platform Updates — July 31–August 7, 2026
13 production image releases (1.120.0 → 1.132.0) plus 39 iblai-web-frontend package releases — 52 in all: admin-set LLM spend caps enforced across tenant, agent, and per-user scopes; the Agent Computer Protocol and in-chat computer use in OS; a grading results API with grader components and a results tab; a new first-party iblai LLM provider covering chat, embeddings, and TTS; SEO foundations across OS and LMS; and notarized macOS desktop builds.
Syracuse Case Study, Expanded: Clementine in Production
The Syracuse University case study now carries the university's own published numbers — 2,038 students and 8,217 questions in 37 days of registration season — plus the classroom use cases Syracuse runs on the platform, and direct links to the four pages ITS publishes at ai.syracuse.edu.
Platform Updates — July 24–31, 2026
10 production image releases (1.109.0 → 1.119.0) plus the iblai-web-frontend packages: per-account usage checkpoints and OpenClaw worker cost attribution in the Data Manager, infinite-scroll projects and SDK-sourced agent settings in OS, nonce-based CSP enforced by default across the SPAs and web-utils v2, a v2 iblai-js and Auth SPA, field-level RBAC write filtering, the v1 application feature, and all-interface network binding in the CLI.
A Course Authoring Agent That Builds Canvas Courses End to End
A new Course Authoring agent takes a course description in plain English and assembles the whole thing in Canvas — shell, pages, assignments, quizzes, modules, module items — in dependency order, unpublished, idempotent, and ready for a human to review before students ever see it.
The ibl.ai Mac App — Offline LLMs and Sandboxed Agents
A native macOS app with everything the web platform does — chat, voice, MCP connectors, any LLM with your own keys — plus two things a browser can't do: local models that keep working with the network off, and agent tool execution sandboxed inside the app.
Give an Agent a Phone Number: Outbound Calls from Your Own Server
A new end-to-end tutorial in the ibl.ai API repo: stand up an agent runtime on your own VPS, teach it to place real outbound phone calls with one skill file and one shell script, and bind it to your organization over REST. About 30 minutes and roughly $4/month of infrastructure.
New Call Interface with Real-Time Transcription
Voice calls with agents get a redesigned in-call interface and a live transcript that streams both sides of the conversation as it happens — then lands in Analytics → Transcripts when the call ends.
Platform Updates — July 22–24, 2026
5 production image releases (1.104.0 → 1.108.0): granular per-object edX course-creation tools, searchable chat history and an Evals tab in OS, a nonce-based CSP, empty-tenant onboarding in the LMS, and Microsoft SSO plus embed-auth hardening in Auth.
ibl.ai Named an OpenAI Select Partner
ibl.ai has been named an OpenAI Select Partner — without lock-in. Already a Google, Microsoft, and AWS partner, the platform stays LLM-agnostic: deploy on your own infrastructure with full source-code ownership and swap models anytime.
Platform Updates — July 16–21, 2026
14 production image releases (1.87.0 → 1.102.0): granular edX course-creation tools, next-intl internationalization, a Chrome side-panel extension, an LMS sidebar and navbar revamp, and an independent second OpenID SSO backend.
Platform Updates — June 2026
74 production image releases (1.25.1 → 1.61.1): agent-driven edX course creation, streaming text-to-speech, first-class privacy-mode chat, the agentAI→OS and Skills→LMS renames, principal-agnostic RBAC, and optional AWS WAF in the infra CLI.
Claw Agents: 48 Pre-Built Agent Configurations Now Available
48 workspace-ready agent configurations for OpenClaw and NemoClaw instances, organized across higher education, enterprise, K-12, and small business verticals.
App CLI: Scaffold Production-Ready AI Applications
Interactive CLI for generating Next.js 15 applications with chat interfaces, authentication, and full ibl.ai SDK integration. Supports AI-assisted customization.
MCP Servers: 7 Model Context Protocol Servers for ibl.ai
A collection of MCP servers covering analytics, agent creation, agent chat, search, user management, and Canvas LMS integration. Connect from Claude Desktop, Cursor, or Claude Code.
Infrastructure CLI: End-to-End AWS Provisioning
Interactive CLI for provisioning the ibl.ai platform on AWS with Terraform and Ansible. Also supports bootstrapping existing servers on any cloud or bare metal.
.iblai Agent File Format: Portable Agent Definitions
A Markdown-based portable agent definition format. One file captures everything needed to define, share, and load an AI agent across tools, teams, and platforms.
Claw Setup: Connect Self-Hosted Agent Servers
Connect self-hosted OpenClaw and NVIDIA NemoClaw servers to the ibl.ai platform. Run your own agent infrastructure with full management through ibl.ai APIs.
Release notes, changelogs, and development activity from across the ibl.ai platform — the applications, the infrastructure tooling, and the open source projects behind them, sourced from GitHub releases, commit logs, and internal development streams. Each release ships to customers running ibl.ai on their own infrastructure: you own all the code and the data, on any cloud, in your VPC, on-premise, or fully air-gapped.