CorvinOS rises to 25/36 as assistants slip
Signal is concentrated today: CorvinOS moved up to 25/36, while Canva Magic Studio fell to 8/36 and Codex CLI fell to 9/36. That points to the same operator read from two directions: a domain specialist is gaining ground, while general-purpose assistant offerings are being re-scored lower as their packaging and scope shift. GitHub Copilot also rewrote its site around cloud agents, MCP servers, and a desktop app, which is a reminder that product surface area is moving fast even when the score doesn’t move the same day. The practical check is whether your workflow needs a specialist or a broader agent platform right now.
Foundation-model news
|
GPT 5.6 · OpenAI
Model release
Foundation-model release — capability changes propagate to every listed tool that uses this model.
Source: Latent Space (swyx + Alessio Fanelli) — https://www.latent.space/p/ainews-openai-launches-gpt-56-solterraluna.
|
|
Grok 4.5 · xAI
Model release
SpaceXAI announced Grok 4.5 as its newest model for software development, reasoning, and productivity work. The company says it was built with Cursor and now runs faster with better token efficiency. For those tracking agentic AI product launches, this shows which models are being tuned for real work, not just chat. The key question is what actually improved enough to push inference to 80 TPS.
Source: MobiGyaan — https://www.mobigyaan.com/spacexai-grok-4-5-update-2026-features.
|
What moved today
|
25/36 · Domain Specialist
Score now 25/36 (22→25). The rubric moved on documented capability evidence; press alone does not move the score.
Trigger: submission.
|
|
8/36 · Guided Assistant
Score re-evaluated after: Introducing Canva Grow 2.0: Create, Launch, and Optimize Ads in One Place
Trigger: MarTech Series — "Introducing Canva Grow 2.0: Create, Launch, and Optimize Ads in One Place".
|
|
9/36 · Guided Assistant
Score re-evaluated after: OpenAI says its Codex workspace agents can run for hours, while Paradigm draws crypto watchers into the mix.
Trigger: Crypto Briefing — "OpenAI says its Codex workspace agents can run for hours, while Paradigm draws crypto watchers into the mix.".
|
|
5/36 · Reactive Tool
Score re-evaluated after: Salesforce Launches MCP-Powered Capabilities for Slackbot
Trigger: CRM Magazine — "Salesforce Launches MCP-Powered Capabilities for Slackbot".
|
|
4/36 · Reactive Tool
Score re-evaluated after: Perplexity Launches Legal AI Platform: 20-Model Agent Targets Law Firm Workflows Above Westlaw
Trigger: Tech Times — "Perplexity Launches Legal AI Platform: 20-Model Agent Targets Law Firm Workflows Above Westlaw".
|
What changed on the ground
Changes we detected on each tool's own pages — pricing, features, platform reach — via our daily crawls. Often ahead of any press coverage; each links to the scored listing.
|
19/36 · Domain Specialist
GitHub Copilot adds agent workflows, MCP servers, and new pricing tiers — GitHub Copilot’s site was rewritten to highlight a broader agentic product suite, including cloud agents, a Copilot desktop app, project knowledge with Spaces, and MCP server support.
Source: the tool's own site — detected 2026-07-10 by agentic.ai's daily recrawl.
|
|
15/36 · Adaptive Collaborator
Claude revamps plans, adds major new features, and teases Fable 5 model tier — Claude’s website now shows updated pricing and plan packaging, including a lower Pro annual price, a new Max starting price, and expanded Free/Pro feature sets.
Source: the tool's own site — detected 2026-07-08 by agentic.ai's daily recrawl.
|
|
9/36 · Guided Assistant
CrePal launches pricing page with Plus, Pro, and Max tiers — CrePal has added a pricing page for the first time, introducing three paid plans with monthly and yearly billing, credit allowances, concurrency limits, commercial-use rights, and add-on credit packages.
Source: the tool's own site — detected 2026-07-10 by agentic.ai's daily recrawl.
|
|
9/36 · Guided Assistant
Manus says it is now part of Meta as it pivots toward business app building — The site now prominently states that Manus is part of Meta, and the page has been rewritten around launching production-ready business applications without engineering.
Source: the tool's own site — detected 2026-07-08 by agentic.ai's daily recrawl.
|
|
17/36 · Adaptive Collaborator
ChatGPT adds a new Go plan and updates pricing/features with GPT-5.5 tiers — ChatGPT’s pricing page was overhauled to introduce a new Go tier and revise the plan matrix across Free, Plus, Pro, Business, and Enterprise.
Source: the tool's own site — detected 2026-07-05 by agentic.ai's daily recrawl.
|
|
18/36 · Adaptive Collaborator
Goose moves to the Agentic AI Foundation at the Linux Foundation — Goose’s page now says the project has moved from block/goose to the Agentic AI Foundation (AAIF) at the Linux Foundation, signaling an ownership and governance transition.
Source: the tool's own site — detected 2026-07-06 by agentic.ai's daily recrawl.
|
From Agentic-News.ai
Agentic-AI industry signal we're tracking — launches, releases, and research from across the ecosystem. The directory rubric hasn't scored these yet; awareness now, evaluation next pass.
|
AgentPrizm
product launch
AgentPrizm launched its AgentMemory and AgentSkills platform on July 9, 2026 for developers and enterprise teams. The product pairs a REST API with MCP infrastructure to give agents persistent memory across sessions. For those tracking agentic AI products, this targets one of the hardest problems in deployment, which is trust in what an agent remembers and reuses. The striking detail is the built-in audit trail that can show each memory decision back to the user.
Via Agentic-News.ai — WBOC TV-16. Read: https://www.wboc.com/online_features/press_releases/agentprizm-launches-governed-ai-agent-memory-platform-that-lets-agents-prove-what-they-remember/article_fbe24eab-69a6-5b1d-b5bc-ce51f5efbdbe.html
|
|
M-Files
product launch
M-Files launched M-Files for Consulting on July 9, 2026, a purpose-built app for consulting firms. The product organizes work from proposal to delivery and adds governed access to confidential client information. For those tracking agentic AI product launches, this shows vendors packaging automation around a specific workflow, not a generic assistant. The key question is how much of the compliance and document grind the new system can actually remove.
Via Agentic-News.ai — FinanzNachrichten.de. Read: https://www.finanznachrichten.de/nachrichten-2026-07/68986470-m-files-launches-m-files-for-consulting-transforming-a-firm-s-knowledge-into-an-ai-powered-advantage-200.htm
|
|
International Telecommunication Union
product launch
The International Telecommunication Union has launched a new Focus Group on Trust and Identity for Humans and Agentic AI. Announced at the AI for Good Global Summit, the effort aims to build frameworks for trusted digital identity and accountable AI behavior across the agent lifecycle. For those tracking agentic AI, this signals that governance is moving closer to the core technical stack, not just the policy layer. The biggest question is whether global standards can keep up with agents that keep getting more autonomous.
Via Agentic-News.ai — Mirage News. Read: https://www.miragenews.com/itu-unveils-global-ai-standards-for-trust-1707644
|
Biggest movers this week
Listings with the largest week-over-week jump in outbound-click interest
- Bolt.new · +3.7% week-over-week. Bolt is an AI coding environment for building websites, apps, and prototypes from a prompt. It also connects to Figma and GitHub so you can start from an existing design or codebase.
Biggest declines this week
Listings with the largest week-over-week drop in outbound-click interest
- Cursor · -36.6% week-over-week. Cursor is a developer-focused AI environment that adds agents, context, and automation around your repositories. It combines an editor-like interface, a CLI, and a cloud agent API to automate code review, bug fixing, CI hygiene, and more. Designed for individual developers and engineering teams who want AI to take real actions in their code and infrastructure, not just chat.
- Claude Code · -16% week-over-week. Claude Code is Anthropic's agentic coding tool that lives in your terminal. It understands your entire codebase, makes multi-file edits, runs commands, manages git workflows, and uses MCP for tool integration. Built with a Unix philosophy — it reads, plans, edits, and verifies in a loop. The fastest-growing product in the coding agent category.
Agenticness Leaderboard
This week's most-clicked tools, re-ranked by rubric score. The score is the editorial signal; the popularity is what drew the directory's traffic here this week.
- Cline · 19/36 · Domain Specialist · L3
Cline is an AI coding assistant for VS Code that can inspect your project, edit files, run terminal commands, and use a browser while asking for permission at each step. It is aimed at developers working on real codebases who want more than code completion.
- Windsurf · 19/36 · Domain Specialist · L3
Windsurf Editor is an AI-powered IDE for developers that blends chat, autocomplete, and agentic code actions into the editor itself. It’s available for Mac, Windows, and Linux, and is aimed at speeding up day-to-day coding work on real codebases.
Wingman is a messaging-first autonomous AI agent by Emergent. It connects to your email, calendar, Slack, CRM, and GitHub, then handles scheduling, research, sales support, and admin tasks through WhatsApp and Telegram conversations.
What operators are searching for this week
A week's demand signal — what operators are typing into the directory's search. Suggestion-chip defaults + off-topic / greetings filtered.
- "AI agent to triage and reply to my email inbox"
- "best coding agent for a large existing codebase"
- "tools to automate customer support ticket triage"
- "AI agent that can browse the web and fill out forms"
- "self-hosted open-source AI agents I can run locally"
Learn: Action Capability: can it actually DO things (run code, send, edit files), or only suggest?
Action Capability asks a simple question: does the tool change the world, or just talk about changing it? A chatbot can draft an email, summarize a file, or suggest code. An agent with action capability can also send that email, edit the file, run the script, or call the API itself. That gap is the difference between advice and work. A clean check is to look for the action surface, not the marketing copy. Does the tool show real connectors, execution logs, write permissions, or a confirmation step before it acts? If all you see is text output, it’s probably a planner, not an actor.
How our 9-dimension rubric works →
One tool worth testing
Today's signal didn't produce a clear single-tool pick worth a test session. Tomorrow's read picks it back up.
Our pick: n8n
The use case: automating my team's repetitive multi-step workflows — lead routing, data sync, and scheduled jobs — without a dedicated engineer.
Tops our independent /36 — strongest on action and reliability, ahead of Zapier. Per-execution (not per-step).
Ranked by our 9-dimension Agenticness rubric — not by who pays us (nobody does). The full scorecard, true-cost pricing, and the reliability evidence behind every score are in the report.
Read the full report →
|