Tadas Antanavicius

375 posts

Tadas Antanavicius

Tadas Antanavicius

@tadasayy

Co-creator of https://t.co/Z9EyZkkWs7, the best way to keep up-to-date with the Model Context Protocol (MCP) ecosystem. tadasant on most other platforms.

United States Katılım Eylül 2010
287 Takip Edilen676 Takipçiler
Tadas Antanavicius
Tadas Antanavicius@tadasayy·
I say this as someone very passionate about the promise of AI-enabled productivity. But clankers that scrape and automate web browsers are not the way. Service provider APIs must be respected. MCP servers let them set the terms of what's meant for agents. That is how we all align
English
0
1
5
206
Tadas Antanavicius
Tadas Antanavicius@tadasayy·
GitHub is being overrun by bots. Reddit just announced "bots must wear name tags". X is several months into a big anti-bot campaign. It's now a real problem - expect all those clankers impersonating humans to be fully blocked from doing anything useful very soon.
English
1
0
7
162
Tadas Antanavicius
Tadas Antanavicius@tadasayy·
MCP is dead? Check the numbers. MCP is booming.
Tadas Antanavicius tweet media
English
2
1
24
2.1K
Tadas Antanavicius retweetledi
Gergely Orosz
Gergely Orosz@GergelyOrosz·
MCPs are the opposite of dead. They are the life blood of how AI agents use services inside mid-sized and above companies. Case in point: Uber runs on MCPs internally, for good reason. Details: newsletter.pragmaticengineer.com/p/how-uber-use…
@levelsio@levelsio

Thank god MCP is dead Just as useless of an idea as LLMs.txt was It's all dumb abstractions that AI doesn't need because AI's are as smart as humans so they can just use what was already there which is APIs

English
92
100
1.3K
265.8K
Tadas Antanavicius retweetledi
Jiquan Ngiam
Jiquan Ngiam@JiquanNgiam·
CLI: Good for single-user mode, hackers. MCP: Good for multi-user mode, teams/enterprises.
English
8
3
35
6.8K
Tadas Antanavicius retweetledi
yenkel
yenkel@yenkel·
if your API has - dynamic client registration via CIMD - user level oauth so agents can auth on their behalf - ability for background agent auth (no UI) - good docs for AI then your API is good enough also, you kind of have an MCP server
Rhys@RhysSullivan

the overindexing on CLIs is kind of insane to me it's building a primitive that's not portable, properly discoverable, has no good approval flow DCR / CIMD to APIs would go so much further but CLI is just the current hype thing

English
3
6
80
12.1K
Tadas Antanavicius retweetledi
David Soria Parra
David Soria Parra@dsp_·
Claude Cowork works because it’s connects to your services via MCP. Your Slack Claude Code plugin works because of MCP. A common auth standard is hugely valuable. If you have context bloat, get a better client, because you shouldn’t just dump a list of tools into your context window.
Rhys@RhysSullivan

MCP sucking is a harness problem, not an MCP problem MCP unlocks behavior that is fundamentally impossible to get via CLI or APIs Bad auth, too much context usage, all get solved with an execution layer - your agent writes code to progressively discover and call tools

English
29
28
215
39.1K
Tadas Antanavicius retweetledi
Rhys
Rhys@RhysSullivan·
MCP sucking is a harness problem, not an MCP problem MCP unlocks behavior that is fundamentally impossible to get via CLI or APIs Bad auth, too much context usage, all get solved with an execution layer - your agent writes code to progressively discover and call tools
Garry Tan@garrytan

MCP sucks honestly It eats too much context window and you have to toggle it on and off and the auth sucks I got sick of Claude in Chrome via MCP and vibe coded a CLI wrapper for Playwright tonight in 30 minutes only for my team to tell me Vercel already did it lmao But it worked 100x better and was like 100LOC as a CLI

English
168
90
1.4K
375.5K
Chris Frantz
Chris Frantz@frantzfries·
everything in the GUI should be available in the API this will be the theme for us for the remainder of the year the UI becomes the preview layer for agentic work if you can click it, you should be able to prompt it
English
33
18
203
13.8K
Gergely Orosz
Gergely Orosz@GergelyOrosz·
@frantzfries gotcha. yeah I hear you I wonder how this will be different... will it be sustainable? I DO remember when I started my first job seniors thought those using autocomplete would honest to god be terrible devs BUT AI allows you to turn your brain off, true (if it's how you do it)
English
6
0
30
2.7K
Gergely Orosz
Gergely Orosz@GergelyOrosz·
Same version of this from ~20 years back: "I swear to god the average new grad is Google-pilled to the point of being unable to complete basic tasks on their own, without searching online" (prevailing wisdom back then was that real devs know stuff by heart or use books)
Chris Frantz@frantzfries

I swear to god the average tech bro is ai pilled to the point of being unable to complete basic tasks on their own anymore insane shit, we are losing the ability to think

English
42
27
769
72.3K
Tadas Antanavicius
Tadas Antanavicius@tadasayy·
The Upwork benchmark made sense as a point-in-time experiment, but surely the goalpost has moved on that, right? Now that coding agents are more widely adopted, I would assume the amount of agent-ifiable Upwork jobs commanding human labor rates has plummeted to near-zero I have no doubt the Devin of today would perform exceptionally well on Upwork tasks that were being posted in 2024, just curious how you might've evolved thinking (or re-inventing) the benchmark here
English
0
0
0
28
Tadas Antanavicius
Tadas Antanavicius@tadasayy·
Every time I try to use an agent to drive a new, fully autonomous workflow, it fails ... but then I break it down, babysit it for a week or two while I catalog all the failure modes in a semi-manual process and it feels like magic on the other side. Don't skip the first steps!
English
0
0
2
69
Tadas Antanavicius
Tadas Antanavicius@tadasayy·
(6/6) The “aha” moments are just the tip of the iceberg. It all turns into a rich, deep craft with the north star of architecting a system to provide the model with sufficient context to produce the working (and mostly good) code. Providing the right context at the right time is a complicated problem. "Write better prompts" is hardly the start of it.
English
1
0
2
74
Tadas Antanavicius
Tadas Antanavicius@tadasayy·
Software engineers worried about AI absorbing their craft would do well to embrace the change and do what the history of building software has always done: move up the abstraction layer. Agentic engineering is here. Teams are making it work. The portfolio of very public examples of this adoption grows by the day. I’ve been coaching some of the teams going through this shift in the software engineering trade - and I’ve found that the mindset shift often takes just two very specific “aha” moments.
English
1
2
6
207
Tadas Antanavicius retweetledi
Jiquan Ngiam
Jiquan Ngiam@JiquanNgiam·
When we ask an agent to "fix this bug," we're approving hundreds of actions we'll never see. These agents have our production credentials; and increasingly, they run in the background while we work on something else. Engineers are spinning up multiple Claude Code sessions in parallel, and ClawdBot hit >100k GitHub stars in weeks by letting agents handle email, workflows, even car purchases. The shift to agents without constant supervision is here. Today we're launching @MintMCP_AI : governance for AI agents.
English
7
8
30
470K

Keşfet

@frantzfries @GergelyOrosz @AnthropicAI @OpenAI @coinbase @tryramp @Spotify @MintMCP_AI