Authority AI Tools Editorial Team
Editorial Team
The Authority AI Tools editorial team maintains this directory using vendor documentation, dated source checks, product changelogs, and clearly identified hands-on observations where available.
Articles by Authority AI Tools Editorial Team
OpenAI Codex mid-May 2026 updates: Appshots, Goal Mode GA, and Codex CLI 0.131 → 0.133
Between May 18 and May 21, 2026, OpenAI shipped Codex CLI 0.131.0, 0.132.0, and 0.133.0 plus the 26.519 app release introducing Appshots, Goal Mode as standard, remote computer use, and plugin sharing for Business.
OpenAI Codex May 2026 updates: Codex for Chrome + Codex CLI 0.130.0
OpenAI’s Codex changelog highlights May 2026 updates including a new Codex Chrome extension and newer Codex CLI releases (0.129.0 and 0.130.0), plus updated enterprise analytics governance documentation.
Cursor May 2026 updates: PR Review, Build in Parallel, Security Review, admin controls
Cursor’s May 2026 changelog highlights a new PR review experience, parallel plan execution with async subagents, Security Review beta, and new admin controls for models and spend.
Cursor 3.3 Release: PR Review, Build in Parallel (async subagents), and new Enterprise admin controls
Cursor 3.3 adds an all-in-one PR review experience, a Build in Parallel mode that runs async subagents, and several enterprise admin upgrades (model access controls, spend limits/alerts, usage analytics, marketplace policies, and security review beta).
Gemini API May 2026 Update: Gemini 3.1 Flash-Lite GA, Interactions API schema migration, multimodal File Search
May 2026 Gemini API release notes: gemini-3.1-flash-lite reaches GA (May 7), the Interactions API schema changes from outputs→steps with a May 20 default cutover (May 6), and File Search adds multimodal image support with gemini-embedding-2 (May 5).
OpenAI Codex CLI 0.128.0 (April 2026): Persisted /goal Workflows, codex update, and Better Plugin + Permission Controls
Codex CLI 0.128.0 adds persisted /goal workflows you can pause/resume, a codex update command, expanded permission profiles, and smoother plugin marketplace workflows — per the official OpenAI Codex changelog.
Claude Platform (May 6, 2026): Managed Agents Multiagent Sessions, Outcomes, and Webhooks in Public Beta
Anthropic added Managed Agents webhooks, moved Multiagent Sessions + Outcomes into public beta, and shipped vault credential background refresh + improved session/event filtering.
OpenAI Updates the Agents SDK (April 15, 2026): Native Sandboxes, MCP Tools, and Codex-Like File Workflows
OpenAI shipped a major Agents SDK update on April 15, 2026, adding native sandbox execution, a model-native harness, MCP-based tool integrations, and workspace primitives like AGENTS.md instructions and Codex-like filesystem tools.
OpenAI Codex Adds Pets (May 1, 2026): A Small UI for Tracking Agent Progress
OpenAI says Codex now has 'Pets' — a small companion UI you can wake with /pet that gives peripheral status about what Codex is doing.
Claude Code Quality Postmortem (April 23, 2026): Effort Defaults Reverted, Idle-Session Bug Fixed, Limits Reset
Anthropic detailed three separate Claude Code changes that degraded perceived quality (effort default, idle-session thinking retention, and verbosity prompting), and confirmed fixes shipped by v2.1.116 along with a subscriber usage-limit reset.
Cursor Team Marketplace (May 1, 2026): First-Party Plugin Controls Without a Repo
Cursor shipped a Team Marketplace update on May 1, 2026: admins can create a team marketplace without connecting a repository, and manage install behavior for first-party plugins (MCP servers, skills, subagents, rules, hooks).
Cursor Security Review (April 30, 2026): Always-on PR Security Reviewer + Scheduled Vulnerability Scanner
Cursor Security Review entered beta on April 30, 2026 for Teams and Enterprise plans, adding always-on security agents: a PR Security Reviewer that leaves inline findings and a scheduled Vulnerability Scanner that can send Slack updates.
Claude Code v2.1.126 (May 1, 2026): claude project purge, gateway-aware /model, OAuth paste login, managed-sandbox security fix
Claude Code v2.1.126 adds claude project purge to delete local state, lists gateway /v1/models in /model when using ANTHROPIC_BASE_URL, lets claude auth login accept pasted OAuth codes for SSH/WSL/containers, expands dangerously-skip-permissions for protected paths, and ships a managed-sandbox security fix.
Gemini API April 2026 Update: Flex/Priority Tiers, Deep Research with MCP, embedding-2 GA, gemini-3.1-flash-tts-preview
April 2026 in the Gemini API: new Flex and Priority inference tiers (Apr 1), gemini-3.1-flash-tts-preview (Apr 15), Deep Research updates with MCP server integration and File Search (Apr 21), gemini-embedding-2 GA (Apr 22), and the gemini-robotics-er-1.5-preview shutdown (Apr 30).
Cursor SDK (April 29, 2026): Build Programmatic Agents with the Same Runtime That Powers Cursor
Cursor launched a TypeScript SDK on April 29, 2026 that exposes the same agent runtime, harness, and models that power the Cursor IDE. Run agents locally or on Cursor's cloud, integrate them into your own apps, and reach for any frontier model behind a single interface.
Claude Code v2.1.120 → v2.1.123 (April 28–29, 2026): claude ultrareview in CI, Windows Without Git Bash, MCP alwaysLoad, plugin prune
Claude Code's late-April 2026 point releases (v2.1.120–v2.1.123) make /ultrareview runnable from CI as a non-interactive subcommand, drop the Git Bash requirement on Windows in favor of PowerShell, add an alwaysLoad option for MCP servers, and ship claude plugin prune for orphaned plugin dependencies.
Cursor 3.2 (April 2026): /multitask Async Subagents, Better Worktrees, Multi-Root Workspaces
Cursor 3.2 adds /multitask async subagents in the Agents Window, improved worktrees, and multi-root workspaces for cross-repo changes. This update also includes interactive canvases and major Cursor CLI QoL upgrades like /debug, /btw, /config, and /statusline.
Warp Goes Open Source (April 28, 2026): AGPL Client, OpenAI as Founding Sponsor, Oz-Driven Contributions
Warp open-sourced the Warp client under AGPL on April 28, 2026 at github.com/warpdotdev/warp, with OpenAI as the founding sponsor, expanded support for Kimi/MiniMax/Qwen plus an 'auto (open)' router, and a new Oz-agent-driven contribution workflow. (updated with changelog-confirmed TOML settings + /feedback details)
Claude Code Week 17 (April 20–24, 2026): /ultrareview Public Preview, Session Recap, Custom Themes, Web Redesign
Claude Code Week 17 opened /ultrareview to public research preview, made session recap a first-class CLI feature, shipped custom themes you can build and ship in plugins, and redesigned Claude Code on the web with a sessions sidebar and drag-and-drop layout.
Claude Code Week 16 (April 13–17, 2026): Opus 4.7, xhigh Effort, Routines, /ultrareview, /usage, Native Binaries
Anthropic's Claude Code Week 16 update made Opus 4.7 the default on Max and Team Premium, added the xhigh effort level and /effort slider, shipped Routines and /ultrareview on the web, a /usage breakdown, and moved the CLI to native binaries.
OpenAI Codex April 2026 Update: GPT-5.5, Browser Use, and CLI 0.124–0.125
What shipped in OpenAI Codex during April 2026: GPT-5.5 as the new frontier coding model, in-app browser use, automatic approval reviews, and Codex CLI 0.124.0/0.125.0.
Claude Code Week 15 (April 6–10, 2026): /ultraplan, Monitor Tool, /autofix-pr, /team-onboarding
Anthropic's Claude Code Week 15 update added /ultraplan cloud planning, the Monitor tool with self-pacing /loop, /autofix-pr from the terminal, and /team-onboarding.
Claude Opus 4.7 (April 16, 2026): /ultrareview, Auto Mode for Max, xhigh Effort Level, and a Better SWE Score
Anthropic released Claude Opus 4.7 on April 16, 2026. Pricing is unchanged at $5/$25 MTok; the model improves SWE-bench-class coding, ships a new /ultrareview slash command in Claude Code, extends auto mode to Max users, and introduces xhigh as the new default effort level.
Google Antigravity 1.22.2 (April 7, 2026): Unified Permissions System for Agent Actions
Google Antigravity 1.22.2 introduces a unified permissions system to control agent actions, improving safety and oversight for agentic workflows.
Showing the latest 24 of 82 articles. Browse the full archive.