Archive

Dive into a collection of past posts and rediscover timeless content.
5 minutes to audit an MCP config

Sep 30, 2026

•

3 min read

5 minutes to audit an MCP config

A safe preflight for the configs, tools and skills your agent can now touch.

Swati Gupta
Swati Gupta
The AI model trained to think less

Sep 30, 2026

•

3 min read

The AI model trained to think less

Plus: an outcome-first prompt, three free ops tools and browser checkout with a buyer gate.

Swati Gupta
Swati Gupta
OpenAI hit pause. Your agents need a brake.

Sep 28, 2026

•

3 min read

OpenAI hit pause. Your agents need a brake.

Plus: Gemini buys on Flipkart, a new scam checker and 3 free building blocks.

Swati Gupta
Swati Gupta
The AI model that refuses to write back

Sep 25, 2026

•

2 min read

The AI model that refuses to write back

Plus: a free prompt to map any messy task into a safe decision flow.

Swati Gupta
Swati Gupta
Jev

Sep 22, 2026

•

4 min read

Jev

Jev: fastest adoption in Vercel history

Swati Gupta
Swati Gupta
Anthropic says slow down. Then Reuters found this

Sep 21, 2026

•

3 min read

Anthropic says slow down. Then Reuters found this

Plus: Claude now builds 26% of Anthropic itself

Swati Gupta
Swati Gupta
OpenAI published six ways its models misbehaved

Sep 18, 2026

•

4 min read

OpenAI published six ways its models misbehaved

One model told its future self to ignore the rules. Here's the audit I run now.

Swati Gupta
Swati Gupta
Your AI provider might be quietly degrading you

Sep 15, 2026

•

4 min read

Your AI provider might be quietly degrading you

The new US advisory tells labs to degrade, not block. Here's how to check yours. →

Swati Gupta
Swati Gupta
The week the labs stopped blaming the plumbing

Sep 11, 2026

•

4 min read

The week the labs stopped blaming the plumbing

Anthropic overturned its own verdict. California signed. Agents kept leaving.

Swati Gupta
Swati Gupta
Load more