Research · everything Intel keeps

Everything Intel keeps, in one place — every brief we run, every department, readers' own, and every finding behind them. Browse below, or see the same body of work as a connected graph on the live map.

The lead

Oct 17, 2026

AI-Assisted Software DevelopmentGitHubnotableOct 15, 2026

GitHub Copilot will drop Claude Opus 4.8, Claude Opus 5 and GPT-5.6 Terra on 16 November

GitHub said on 15 October 2026 that it will deprecate Claude Opus 4.8, Claude Opus 5 and GPT-5.6 Terra across all GitHub Copilot experiences, including Chat, inline edits, ask and agent modes and code completions, on 16 November 2026.

What it means for your work

Workflows and integrations pinned to these models need to move before 16 November. Organisations that turned off default model enablement have to enable the replacements themselves.

△ vendor-claimed · github.blog · added todayRead it at github.blog

What today means

Our read on the items that move something. The reporting is everyone's; this part is ours.

This is real production routing data, not a benchmark. Teams are switching labs often and sending most volume to cheaper open models, while frontier models keep a large share of spend.Open-weight models now carry 71.5% of Vercel AI Gateway tokens and 23% of spendThe AI Product & Business Strategy beat · Vercel
Revenue per compute dollar and margins are the numbers investors will use to judge whether frontier-model economics hold up. These are anonymously sourced figures given to investors, not audited results.Anthropic told investors quarterly revenue rose nearly 50% to $16.8bn, the FT reportsThe AI Product & Business Strategy beat · Simon Willison (citing the Financial Times)
Coding platforms are among the largest model buyers. A public switch on cost grounds shows pricing pressure on frontier labs.Replit is no longer working with Anthropic, its CEO saysThe AI-Assisted Software Development beat · TechCrunch

Share the archiveX LinkedIn Email

Following opens with early access — sign up for early access to register your interest; it does not switch following on today. Everything you follow will land on one page, in your account menu.

All coverage

Everything Intel has read, newest first. Each title opens at its original publisher.

Agent Frameworks & Orchestration
The share of 'Forbidden' responses served to AI agents has roughly doubled since the summer
Agents that shop or browse on a user's behalf will hit more blocks, which pushes builders toward sanctioned APIs and merchant partnerships.
a16z2026-10-16
Agent Frameworks & Orchestration
Autodesk pitches 'Project Intelligence', an AI assistant that turns stated intent into design work
If the main design-software vendor moves from tools to an assistant that does the work, how practices staff projects and check deliverables will change.
AEC Magazine2026-10-16
Agent Frameworks & Orchestration
Google ADK 2.12 cuts long-session latency by 40% and adds human sign-off before workflow steps
Filtering tool output before the model reads it is a direct defence against injected instructions, and per-step approval makes risky workflows easier to put in production.
Google2026-10-13
Agent Frameworks & Orchestration
Zapier folds Zapier Agents into a single 'AI by Zapier' step
Teams with Zapier Agents in production should use the feature mapping to move them, since agent behaviour now sits inside regular Zaps.
Zapier2026-10-06
Agent Frameworks & Orchestration
Anthropic's new Cowork runs its sandbox VM in the cloud instead of on your laptop
Sessions now run in Anthropic's cloud, with the desktop app passing local files to them, so any data-handling review done for the old on-device Cowork needs redoing.
Simon Willison (quoting Anthropic)2026-10-05
Agent Frameworks & Orchestration
Microsoft Agent Framework 1.20 for Python adds Foundry workflow hosting and new vector-store connectors
Teams already on SQL Server or DuckDB can now use them as the agent's vector store, with no separate vector database.
Microsoft2026-10-02
Agent Frameworks & Orchestration
Google's ADK 2.11 adds graceful cancellation and human approval for workflow tool calls
Being able to stop a run cleanly and requiring approval before a tool call are two controls production agent workflows usually have to build themselves.
Google (Agent Development Kit)2026-10-02
Agent Frameworks & Orchestration
NVIDIA launches an Open Agent Safety Platform, pairing an open-source runtime with a hardware watchdog
Agent containment is moving into the infrastructure layer; teams deploying autonomous agents should watch whether OpenShell becomes a de facto runtime boundary.
NVIDIA2026-09-28
Agent Frameworks & Orchestration
n8n launches Agents that sit alongside its workflows and can call them as tools
Teams already running n8n can try open-ended agents while keeping hard limits by exposing only vetted workflows as tools.
n8n2026-09-25
Agent Frameworks & Orchestration
Claude Agent SDK adds a verbatim_prompts option to stop untrusted text triggering file reads or commands
If your agent puts fetched web pages, emails or user text into prompts, turning this on removes an easy injection route.
Anthropic2026-09-23
Agent Frameworks & Orchestration
A practical pattern for giving coding agents a memory you control
Agent memory is the difference between re-explaining your codebase every session and not — and owning it keeps it portable across tools.
Hugging Face2026-09-03
Agent Frameworks & Orchestration
WeChat Pay opens its agent payment card to DeepSeek Harness and OpenClaw
Agent-initiated payment is shipping at consumer scale with a segregated balance and a human confirmation step — a concrete answer to "how do you let an agent spend money".
TechNode2026-09-01
Agent Frameworks & Orchestration
A working checklist for agents that have to survive longer than one call
The single line worth stealing: write down the done condition before the agent starts, and check it with something deterministic rather than another model.
n8n2026-08-31
Agent Frameworks & Orchestration✓ verified
OpenClaw 2.0 rebuilds its control interface and moves sessions into SQLite
Two upgrade traps in one release: downgrading now means a manual SQLite restore, and shared sessions are not a permission boundary you can lean on.
OpenClaw2026-08-30
Agent Frameworks & Orchestration
Vercel lets you scaffold, deploy and chat with an agent from the dashboard
The distance from "we should try an agent for this" to a deployed, Slack-reachable one that calls your own MCP servers is now a dashboard form.
Vercel2026-08-28
Agent Frameworks & Orchestration✓ verified
Anthropic opens a research preview of a shared standard for agents to drive lab and factory hardware
The tool-calling layer that made software agents useful is being pointed at instruments and machines; if it holds, the same harness that drives an API drives a microscope or a robot arm.
Anthropic2026-08-27
Agent Frameworks & Orchestration
Microsoft's Agent Framework gives.NET builders a first-class way to intercept an agent mid-run
A supported interception point is the difference between auditing what an agent did and only reading about it afterwards.
Microsoft2026-08-22
Agent Frameworks & Orchestration
Nvidia research puts the agent harness, not the model, at the centre of the result
If the harness carries the outcome, agent quality is engineering you own rather than a model you rent, and it is the half you can actually iterate on.
TechCrunch2026-08-21
Agent Frameworks & Orchestration✓ verified
Nvidia's agent wrapper took Claude Opus 5 from 30% to 100% on a reasoning benchmark
If the software around the model — memory, supervision, tools — is worth seventy points on a long task, then agent quality is mostly engineering you control rather than a model you wait for.
NVIDIA2026-08-21
Agent Frameworks & Orchestration
Microsoft's Agent Framework adds agent-generated interfaces and a fatal-middleware signal
An agent that renders its own interface, and a middleware failure that actually stops the run, are both about giving the operator somewhere to stand.
Microsoft2026-08-21
Agent Frameworks & Orchestration✓ verified
n8n adds an Amazon Bedrock AgentCore node for multi-agent teams with persistent customer memory
Persistent per-customer memory is the feature that turns a workflow automation into something that behaves like an account team — and the fact that it now arrives as a node rather than a build is what puts it in reach of a non-engineer operator.
n8n2026-08-20
Agent Frameworks & Orchestration✓ verified
Warp packages the "software factory" as infrastructure teams can rent
A named automation share from an operator — 30-35% of weekly tasks — is a far more useful planning number than a benchmark score, and the phase decomposition it sells is a decent template even for a team that builds its own.
TechCrunch2026-08-18
Agent Frameworks & Orchestration
LangSmith adds evaluators you tune to your own labels, starting with Perceived Error
Tuning the judge on your own labels is the honest version of LLM-as-judge; it also makes the evaluator a maintained asset with a drift problem, which is a cost worth budgeting before adopting one.
LangChain2026-08-18
Agent Frameworks & Orchestration
Cline joins the AI SDK harness layer via an official adapter
A standard harness interface is how agent choice stops being an architecture decision and becomes a dependency you can swap.
Vercel Changelog2026-08-18