AI Insider archive
The daily brief on agentic AI tools, launches, and what changed. Browse past issues below.
· v2-fresh-212
Yutori Delegate turns web-task autonomy into a recovery check
Coarena’s new safety API gives delegated-work teams another way to test risk.
· v2-fresh-210
AutoGen maintenance creates a Microsoft migration check
Kimi's expanded API menu adds a separate model-selection check for builders.
· v2-fresh-207
StackAI's Asana deal makes workflow agents a stack review
MimikFlow opens WhatsApp, email follow-ups, and MCP; repetitive-task demand leads.
· v2-fresh-205
ChatterMate 14→17 makes collaboration a review item
Collate 2.0 adds governed agents; Clay now separates Actions from Data Credits.
· v2-fresh-203
Replit's pricing shift changes your coding-agent cost check
OpenHands now puts isolation, PR creation, and audit logs in one production frame
· v2-fresh-201
Devin Desktop changes your coding-agent control plane
BrowserAct adds reusable skills and persistent sessions for web workflows.
· v2-fresh-199
Reply.io's 70-tool MCP makes sales-agent access a check
The key question is which outbound actions you will let an assistant execute.
· v2-fresh-197
Semantic Kernel users now need a 1.0 migration plan
Microsoft names an enterprise-ready successor with A2A and multi-provider support.
· v2-fresh-195
AutoGen maintenance mode makes your framework exit plan urgent
Cursor leads outbound clicks; GitHub Copilot interest is up 42.3% week over week.
· v2-fresh-192
Claude Code changes your individual-plan calculus
Whole-repo coding demand is rising as vendors widen the scope from access to delivery.
· v2-fresh-190
Devin Desktop makes coding-agent fleet control a buying check
Cline adds a native app as whole-repo work drives this week’s only search demand.
· v2-fresh-188
BrowserAct Data API makes browser agents a build-or-buy check
Writer brings Playbooks into Slack while Kimi K3 claims 1M-token context.
· v2-fresh-185
AutoGen’s maintenance notice puts your agent stack on notice
The replacement question now spans memory portability and governed tool access.
· v2-fresh-182
Claude’s plan revamp turns MCP into a buying check
New Education tier and Microsoft 365 support change who can justify it.
· v2-fresh-180
Coda just became a docs migration decision
Superhuman Docs adds AI Views and enterprise databases while Copilot shifts the coding surface.
· v2-fresh-178
Cursor’s pricing rewrite changes your agent stack math
Seat types split into Standard, Premium, and Free; on-demand usage is now spelled out.
· v2-fresh-176
Buffer’s Team plan drops to $60 billed yearly
Runway also widened model access and added workflow/MCP surfaces.
· v2-fresh-174
Semantic Kernel just became a migration decision
AutoGen’s maintenance mode makes Microsoft Agent Framework the other half of the read.
· v2-fresh-172
Windsurf becomes Devin Desktop in your migration decision
Plans and settings carry over; Beautiful.ai now hands Claude a verified connector.
· v2-fresh-170
MimikFlow’s MCP beta puts inbox actions in your eval queue
Sana moved inside Workday, and operators are still asking for tools that can act, not chat.
· v2-fresh-168
Claude just became a team rollout decision
Cursor leads outbound clicks this week as repo-wide coding-agent searches stay hot.
· v2-fresh-166
GitHub Copilot just became a workspace decision
Clay rewrote pricing into usage-based tiers; Moveworks is now ServiceNow-branded
· v2-fresh-164
Chrome’s Gemini browse lands in your eval queue
AutoGen goes into maintenance mode, and UiPath posts a new pricing page for Automation Hub.
· v2-fresh-162
Framer AI just became a page-ops decision
Inworld cuts prices while driftless jumps to 21/36 on submission
· v2-fresh-160
Semantic Kernel just became a migration decision
LangChain is widening LangSmith while AI Interviews cuts unit pricing
· v2-fresh-158
Kilo Code turns coding help into an enterprise control decision
Notion AI’s new Business tier and coding-agent demand point to the same check: control.
· v2-fresh-156
Canva AI 2.0 turns creative work into an agent decision
Composio’s new usage-based tiers make the control plane part of the bill.
· v2-fresh-154
BrowserAct just turned browser agents into an infra bill
Devin widened its product suite, while AutoGen moved to maintenance mode.
· v2-fresh-152
ChatGPT just turned pricing into a plan-choice problem
The new Go tier is part of a broader pricing rewrite; Replit Agent moved too.
· v2-fresh-150
Claude Code just changed your plan math
Remote MCP now sits inside Anthropic’s revised plans; check what moved with it.
· v2-fresh-148
Semantic Kernel owners now have a migration check
Chrome’s new auto-browse mode widens the gap between chat and work done.
· v2-fresh-146
Windsurf’s rename turns Devin Desktop into a migration decision
The new page now frames it around fleets of local and cloud agents, not just a desktop app.
· v2-fresh-144
Cursor’s platform expansion changes your build stack
Origin, headless CLI, and expanded MCP support shift the test from coding assistant to automation layer.
· v2-fresh-142
Runway’s pricing reset changes your AI video budget
LangChain shipped managed agents with A2A/MCP support; Replit added custom MCP endpoints.
· v2-fresh-140
StackAI’s Asana acquisition changes your enterprise agent stack
Microsoft’s Semantic Kernel is now Agent Framework; ChatGPT adds an $8 Go tier.
· v2-fresh-138
Canva AI 2.0 just landed in your content workflow
Slack, Gmail, Drive, and Notion can now feed editable generations.
· v2-fresh-136
AutoGen just became a migration decision
Glean adds headless MCP support; Perplexity Computer leans into enterprise work
· v2-fresh-134
GitHub Copilot just changed the coding-agent budget
New Free, Pro, Pro+, and Max tiers land as 'best coding agent' searches keep surfacing.
· v2-fresh-132
Semantic Kernel is being superseded; Replit changed the math
Microsoft Agent Framework is the successor; Chrome adds auto browse.
· v2-fresh-130
Zapier’s AI orchestration pricing changes your build math
AI steps, code, and SDK now share one task-based model; Clay adds MCP and agent plugin support.
· v2-fresh-128
AutoGen just became a migration decision
Claude Code flips Auto mode on by default for paid users; ChatGPT adds an $8 Go tier.
· v2-fresh-127
Browser Use lands in your MCP stack
The new CLI server brings persistent Python execution and screenshots into one workflow
· v2-fresh-125
Semantic Kernel becomes Microsoft Agent Framework
Julius AI adds Browser Agent; coding-agent demand still leads search
· v2-fresh-123
v0 launches Business plan at $100/user/month
Suno resets Pro to $8 and Premier to $24; Perplexity adds enterprise controls.
· v2-fresh-121
GitHub Copilot repositions as a broader agentic dev platform
Claude Code now sits in all paid Claude plans; coding-ai demand stays present
· v2-fresh-119
Augment Code rolls out Prism, Code Review, and model updates
Kilo Code changes pricing too, while Cursor and Cline keep leading outbound clicks
· v2-fresh-117
Cursor expands CLI, SDK, and MCP support
Claude Code rises +86.4% while ChatGPT rewrites plan features around GPT-5.6
· v2-fresh-115
Perplexity Computer adds Microsoft 365 support
Goose broadens beyond coding; Cursor still leads outbound clicks
· v2-fresh-113
OpenHands launches Agent Canvas for team automations
GitHub Copilot rewrites for agentic coding; Claude Code rises +63.6%
· v2-fresh-111
Claude Code enters Team & Enterprise tiers
Kilo Code widens into a multi-surface agent platform; Otter.ai adds MCP and calendar links
· v2-fresh-109
Cursor expands Cloud Agent handoff as Claude resets pricing
GPT-5.6 lands on ChatGPT plans; Manus also repositions around Meta and enterprise use
· v2-fresh-107
Cline adds MCP and CLI fixes for agentic coding
Cursor gains +43.2% week over week; Perplexity Computer adds RBAC and credential vault controls
· v2-fresh-105
GitHub Copilot adds native MCP integration
Cline leads outbound clicks this week; “best coding agent” drew 4 searches
· v2-fresh-103
Claude Code adds Free to Enterprise pricing
Kilo Code adds free individual use; Cursor rises +44.1% WoW
· v2-fresh-101
Claude overhauls plans as ChatGPT adds Go
Manus now points to Meta; Cursor and Lindy also reset their pages
· v2-fresh-099
OpenHands launches Agent Canvas as Cline broadens the platform
Agent Canvas goes live with desktop, cloud, and workflow automations; Cursor interest also rose 45.2%
· v2-fresh-098
Claude Opus 5 lands as Perplexity Computer adds publishing
Salesforce Agentforce rises to 9/36; Jasper and C3 AI also moved on site
· v2-fresh-095
GitHub Copilot overhauls pricing and agent plans
Kilo Code adds enterprise controls; Augment ships new MCP connectors
· v2-fresh-093
Cursor revises pricing as Claude and ChatGPT reprice
Manus now says it is part of Meta; plan pages are shifting fast
· v2-fresh-090
Cline deprecates .clineignore as agent governance hardens
Perplexity Computer adds Mac workflows; Jasper broadens into GEO and pipelines
· v2-fresh-089
GitHub Copilot adds custom agents and MCP servers
Augment Code and Kilo Code also widen their platform scope; v0 drops 28%
· v2-fresh-087
ChatGPT unveils Go as Claude and Cursor rewrite plans
Claude Code enters Pro; Cursor widens ACP/MCP support
· v2-fresh-085
Lindy 3.0 adds team controls and 500+ actions
Canva Magic Studio slips to 7/36 as AI-build reach expands
· v2-fresh-083
Motion updates credit pricing as Jasper broadens platform scope
Perplexity adds Comet iOS and multimodal Deep Research; C3 AI pushes MCP-backed integrations
· v2-fresh-082
CrePal, Augment, and Otter add harder edges
Public pricing, MCP exposure, and new integrations are the signal to watch.
· v2-fresh-080
GitHub Copilot adds a Free tier and AI Credits
Cursor ships Cloud Agent handoff; Manus says it is now part of Meta
· v2-fresh-077
Lindy 3.0 and Jasper widen the agent stack
Cursor still edges up; Windsurf drops 32.4% week over week
· v2-fresh-075
Perplexity Computer adds Slack-to-Linear workflows
Augment Code changes pricing; Otter.ai broadens integrations across core systems
· v2-fresh-073
CorvinOS rises to 25/36 as assistants slip
Codex CLI and Canva both moved down; GitHub Copilot broadened its agent suite.
· v2-fresh-063
Amazon Q Developer rises to 10/36 as Amazon leans into agents
Alexa Routines slips, Agentforce moves up, and Canva gains on ad workflows
· v2-fresh-059
Cofounder climbs to 13/36 as GPT-5.6 lands
Pipedream and Make arrive as workflow automation gets fresh options
· v2-fresh-051
Claude up to 16/36 as Opus 4.8 adds parallel subagents
ChatGPT drops to 16/36 while enterprise control concerns stay front and center
· v2-fresh-046
CrowdStrike Charlotte AI jumps 3→8 as agent security firms up
Framer and SentinelOne also move up while Cursor leads outbound clicks
· v2-fresh-038
Gemini climbs to 13/36 as Copilot slips to 18/36
BrowserAct debuts for managed web workflows; Claude Code demand rose +69.6%