Your Claude Code config is now a security surface. Here's a free auditor.
A few days ago I read a post from techriot.io about Anthropic's enterprise Claude Code playbook. The engineering content was excellent. But the post made one observation that stuck with me: when an engineering team adopts Claude Code, they see six new productivity features. The security team inherit
How Narratr exposes its brand intelligence tools to ChatGPT (as a GPT Action) and to Claude.ai (via MCP over HTTP).
Architecture overview Both integrations share the same OAuth 2.0 server and the same 7 tool routes under /api/mcp/tools/ . The difference is how each client authenticates: ClientAuth methodClient secret?DiscoveryChatGPT GPT ActionAuthorization code + client_secretYes ( MCP_GPT_CLIENT_ID / MCP_GPT_CL
From $2,400 to $680: Real Patterns for Claude Cost Control
If you've spent any time with Claude Code or the Anthropic API in 2026, you've probably had this experience: you start what feels like a simple task, the agent starts "exploring," twenty minutes later it's compacting context for the third time, and your monthly subscription is gone before you've shi
What Narratr is becoming — and why it matters now
Most AI content tools have the same problem. They generate fast. They sound fine. And then someone on your team reads it and says — "we would never say it like that" , or worse, "that's not even true." The speed is real. The quality gap is also real. And as AI-generated content floods every channel,
Why Human-Verified AI Brand Analysis Is Your Competitive Edge
The Counter-Narrative Brands Need to Hear Right Now Some platforms are actively marketing against human oversight in brand analysis — framing it as friction, as a bottleneck, as something to be engineered away. That framing is spreading fast. And if it solidifies unchallenged, it will cost brands mo
The Hidden Cost of Fast AI: What We Learned Evaluating Search-Augmented APIs for Competitive Intelligence
Speed is seductive. When one API returns results in 24 seconds and another takes 88, the faster one feels like the obvious choice — especially when the price tag is roughly the same. But when you're generating competitive market intelligence for real brands, speed without accuracy isn't a fe
Twelve Steps to a Cheaper, Better LLM App
*Post 4 of 4 in the **Building With LLMs** series. [See all posts](blog-series-index.md). Previous post: [Tokens and Temperature, in Plain English](blog-series-3-tokens-and-temperature.md)* --- We just finished an end-to-end audit of an LLM-powered web app — 23 prompts, 21 distinct AI calls, a mix o
We don't just audit AEO. We ship it.
The AEO tool category has a tell. Open any of them — HubSpot AEO, AEO Engine, SE Ranking's AEO module, Webflow's new AEO product — and you'll find the same shape of output. A score. A list of recommendations. An estimated traffic lift if you make the changes. Maybe a chart showing which AI engines m
Most AI marketing tools have a missing layer. We built it.
A few weeks ago I watched a brand owner I respect open a fresh ChatGPT tab, paste in their company values, paste in three product descriptions, paste in their three top competitors, paste in the angle they wanted, and ask the model for a LinkedIn post. They had done this six times that week. Each po
Tokens and Temperature, in Plain English
Most of the LLM cost and quality issues I've seen in production trace back to two things people don't intuit about how these models work. One is *tokens* — the unit of measurement everyone gets billed in but few think about in human terms. The other is *temperature* — the single setting that control
Buy vs. Build After AI: What's Still Worth Owning - Part 2
*Post 2 of 4 in the **Building With LLMs** series. [See all posts](blog-series-index.md). Previous post: [Should You Build a SaaS in 2026?](blog-series-1-should-you-build.md)* --- The first post in this series asked the seller's question: should you build a SaaS at all in a world where AI generates
Should You Build a SaaS in 2026? - Part 1
Post 1 of 4 in the * Building With LLMs * series. A quick note on scale before we go further. The app this series is based on is a small LLM app — twenty-three prompts, twenty-one distinct AI calls, run by a small team. The numbers and anecdotes throughout reflect that. But everything below works as
Building With LLMs — A 4-Post Series
A field guide for the rest of us. Written after auditing a real LLM-powered SaaS end-to-end — 23 prompts, 21 distinct AI calls, a mix of providers and fallbacks — and writing down what we learned. The four posts 1. Should You Build a SaaS in 2026? The existential question first. AI generates code in