https://xitter.cf/t.co/WIInSHSiO1 Founder
Joined October 2008
- Tweets373
- Following1.6K
- Followers1.5K
- Likes2.7K
Alex Patterson is working on @CodingCatDev retweeted
Shipping an MCP knowledge pipeline to production requires three guarantees: low token overhead, source drift protection, and strict read-only auth. Never let unvetted doc edits silently mutate agent answers. Production patterns in the reply 👇
Alex Patterson is working on @CodingCatDev retweeted
What happens when someone edits a pricing or policy doc with an unverified claim? In standard RAG, the vector index ingests it immediately. With Sanity Knowledge Bases, the conflict is flagged before it touches agent state. Demo in the reply 👇
Alex Patterson is working on @CodingCatDev retweeted
How do you keep AI agent context windows lean at runtime? Shape the dataset at build time. Use a narrow GROQ projection to strip UI metadata and compile a compact dataset outline so agents retrieve exact entry paths on demand. Details in the reply 👇
Alex Patterson is working on @CodingCatDev retweeted
Anthropic has filed an S-1 registration statement with the SEC related to its proposed IPO.
Read more: anthropic.com/news/s1-sec
Legal disclaimer: A registration statement relating to these securities has been filed with the SEC but has not yet become effective. These securities may not be sold, nor may offers to buy be accepted, prior to the time the registration statement becomes effective. This post does not constitute an offer to sell or the solicitation of an offer to buy these securities, nor shall there be any sale of these securities in any state or jurisdiction in which such offer, solicitation or sale would be unlawful prior to registration or qualification under the securities laws of any such state or jurisdiction.
Alex Patterson is working on @CodingCatDev retweeted
Stop Wasting Tokens: Give AI Agents a Dataset Outline. Keep AI agents accurate and token-efficient by replacing broad similarity searches with a lightweight dataset outline and direct entry-path retrieval. Start with a clear purpose and a narrow Groq proj
Alex Patterson is working on @CodingCatDev retweeted
Broad vector similarity searches burn tokens on noisy chunks and miss structured context. Instead of dumping raw RAG hits into the prompt, give AI agents a lightweight dataset outline with direct entry-path retrieval. Full breakdown in the reply 👇
Alex Patterson is working on @CodingCatDev retweeted
Documentation never stays static. Instead of rebuilding your entire knowledge index blindly, Sanity Knowledge Base runs scheduled weekly or monthly refreshes, diffs sources against your last build, and flags material changes. See it in the reply 👇
Alex Patterson is working on @CodingCatDev retweeted
Why pay token latency at runtime to reconcile conflicting RAG chunks? Moving knowledge compilation to build time with deterministic GROQ projections and Context MCP guarantees your agent reads verified state. Architecture in the reply 👇
Alex Patterson is working on @CodingCatDev retweeted
Traditional vector RAG dumps contradictory chunks into your prompt and expects the LLM to guess which one is current. Compile verified knowledge at build time instead of debugging hallucinations at query time. Breakdown in the reply 👇
Alex Patterson is working on @CodingCatDev retweeted
How do you give an agent access to PDFs, spreadsheets, and Portable Text without paying the token tax on every turn? Hold a lightweight topic outline in `initial_context` and fetch specific entries on demand. Architecture in the reply 👇
Alex Patterson is working on @CodingCatDev retweeted
What happens when you feed an AI Knowledge Base a Sanity dataset, crawled GitHub pages, and a custom lore file with overlapping facts? The ahead-of-time build flags the conflict side by side before your agent ever sees it. Live demo in the reply 👇
Alex Patterson is working on @CodingCatDev retweeted
Sanity Knowledge Base: Lower Token Costs, Higher Accuracy. Unify text, PDFs, spreadsheets, and crawled documentation in Sanity’s topic tree. Fetch only the entries each question needs to reduce token costs and improve accuracy, with scheduled refreshes th
Alex Patterson is working on @CodingCatDev retweeted
Stuffing entire PDFs, spreadsheets, and docs into an LLM prompt burns token budgets and degrades accuracy. Sanity Knowledge Base unifies disparate sources into a single topic tree so agents only load what a query needs. Breakdown in the reply 👇
Alex Patterson is working on @CodingCatDev retweeted
Similarity search is a leaky abstraction for structured knowledge. Instead of hoping top-k vector chunks have the right context, give your agent a purpose-steered outline via `initial_context` and read exact Markdown paths. Architecture in the reply 👇
Alex Patterson is working on @CodingCatDev retweeted
Naive vector RAG chunks raw documents and forces your LLM to resolve contradictions at query time. Hardcoding Markdown in system prompts breaks when content updates. Ahead-of-time Knowledge Base compilation fixes both. Full walkthrough in the reply 👇
Alex Patterson is working on @CodingCatDev retweeted
Today WorkOS is launching auth.md
An open protocol for agents to register for services on the web.
We're partnering with @Cloudflare and @Firecrawl as some of the first providers.
Why did we build this? And why now? 🧵
Alex Patterson is working on @CodingCatDev retweeted
Most support bots have no context about you.
No account, no project, no history. You ask a question, get a generic answer.
We wanted ours to be better.
Episode 2 of Sanity on Sanity goes behind the scenes of building the agent on sanity.io/help with Support Operations Manager Elisa and Support Engineer Darius.
Tune in to see how the setup works, how answers are personalized based on your plan, why the whole prompt and config live in editable content, and how they use a metrics dashboard to catch weak answers and fix them. Then bring your questions for Elisa and Darius in the live Q&A after the episode.
Already in our server? It's under Events, hit Interested for a reminder.
New to Discord? Click Join Server, then Events → Interested.
Discord event link 👉 go.sanity.io/eqKC4rq
Join our Server 👉 go.sanity.io/1Yg0VcD
Alex Patterson is working on @CodingCatDev retweeted
Sanity Studio v6 ships this week, June 11. The pre-release is out now.
What's in it:
→ 2–9× faster builds across the studios we've measured (@vite_js 8, smaller bundles)
→ Faster default search: groq2024 (wildcards, phrases, negation, full Portable Text)
→ React strict mode on by default in dev, which prevents you from shipping bugs
Schemas, config, and content APIs are unchanged. Official plugin support is rolling out.
Two breaking changes to check:
→ Node 20 is now end-of-life and hence dropped, minimum is now 22.12
→ Custom auth providers now replace the built-ins instead of appending (making the experience less confusing)
Test today: npm install sanity@next-major
From June 11: npm install sanity@latest
Upgrade guide, including plugin notes, and the path from earlier versions: sanity.io/blog/sanity-studio…
Alex Patterson is working on @CodingCatDev retweeted
Shannon Elizabeth is coming to Google Next. I know, bold choice.
On Thursday, April 23rd from 3:30-5:30pm, Shannon Elizabeth will be at FusionAuths Mom’s Basement Activation Alley Space, #SAA2 2nd floor outside the expo, for a meet-and-greet, photo opp, and–becasue we fully committed to the bit–handing out signed Raspberry Pis.
Yes, really.
We are turning our activation space into a full-on 90s basement throwback, complete with awkward family photos, Tamagotchis on the spin-to-win, retro games, and the kind of nostalgia that feels deeply family to anyone who has ever built something in a room with bad lighting and excellent ideas.
If you’re heading to Next, come for Shannon Elizabeth and stay to meet the FusionAuth team and see what flexible, developer-friendly CIAM actually looks like.
#googlenext #ciam #customeridentity #fusionauth
Alex Patterson is working on @CodingCatDev retweeted
1M users on Sanity.
The interesting thing is what the content teams among them are doing now.
Content operations used to mean editing pieces one at a time and manually tracking how a change rippled across locales, references, and related docs.
Now teams are running agents against the whole content layer.
A price change propagates across eleven country sites in one pass. An archive of ten years of articles gets audited for coverage gaps in an afternoon. A migration that would take significant engineering work lands overnight.
Content ops has become a system you query, not a queue you work through. That's what the teams shipping on Sanity are doing.
Thanks for building with us!
ALT One million users on Sanity