Claude Code 2.1.172 has been released.
30 CLI changes
Highlights:
• Sub-agents can spawn nested sub-agents up to five levels, enabling deeper multi-agent workflows
• 1M-context sessions without usage credits now auto-compact to standard limits, avoiding permanent stalls
• Marketplace plugin browser (/plugin) adds a search bar to speed plugin discovery and reduce browse time
Full details available in thread ↓
Function calling & MCP for LLMs, explained visually!
Before MCPs became popular, AI workflows relied on traditional Function Calling for tool access. Now, MCP is standardizing it for Agents/LLMs.
The visual below explains how Function Calling and MCP work under the hood.
Today, let's learn:
- Function calling by building custom tools for Agents.
- How MCPs help by building a local MCP client with mcp-use and using tools from the Browserbase MCP server.
In Function Calling:
- The LLM receives a prompt.
- The LLM decides the tool.
- The programmer implements a procedure to accept a tool call request from the LLM and prepare a function call. The tool call request is found in the LLM's response when you prompt it.
- A backend service executes the tool.
This Function Calling takes place within our stack:
- We host the tool.
- We implement a logic to determine the tool to invoke and its parameters.
- We execute it.
So, Function Calling requires us to wire everything manually.
MCP simplifies this!
Instead of hard-wiring tools, MCP:
- Standardizes defining, hosting, and exposing tools.
- Makes it easy to discover tools, understand schemas, and use them.
- Demands approval before invoking them.
- Detaches implementation from consumption.
For instance, whenever you integrate an MCP server, you never write a line of Python code to integrate the tools.
Instead, you just integrate the MCP server and everything beyond this follows a standard protocol handled by the MCP client and the LLM:
- They identify the MCP tool.
- They prepare the input argument.
- They invoke the tool.
- They use the tool’s output to generate a response.
Everything happens through a standard (but abstracted) protocol.
So here’s the key point: MCP and Function Calling are not in conflict. They’re two sides of the same workflow.
- Function Calling helps an LLM decide what it wants to do.
- MCP ensures that tools are reliably available, discoverable, and executable, without you needing to custom-integrate everything.
For example, an agent might say, “I need to search the web,” using function calling.
That request can be routed through MCP to select from available web search tools, invoke the correct one, and return the result.
Check the workflow in the diagram below.
To build a local MCP client, you can use mcp-use because it lets you connect any LLM to MCP servers & build private MCP clients, unlike Claude/Cursor.
- Compatible with Ollama & LangChain
- Render MCP outputs as UI elements
- Stream Agent output async
- Built-in debugging mode, etc
I have shared the mcp-use GitHub repo in the replies!
¡NOTICIÓN! Microsoft usará el kernel de Linux.
Windows 11 migrará progresivamente para mejorar rendimiento y compatibilidad con el cloud.
✓ WSL pasa a ser el sistema base
✓ Compatible con drivers y software actuales
Primera beta - 28 de diciembre de 2026
Shipmas Day 3. Part 2. 🚢🎁
We named our product Stitch for a reason. 🧵
And today, we finally let you do it.
Introducing Prototypes.
You can now select multiple screens and "stitch" them together into a fully functional, clickable user flow.
✅ Create clickable user flows
✅ Test interactions and animations
✅ Ask for edits by clicking divs on the screen
✅ Export the full context to AI Studio or other coding agents
This is Day 1 of interaction design in Stitch, and we have so much more coming.
And yes, you can export your entire prototype flow directly to coding agents like AI Studio.
We are moving from "Generating UI" to "Designing User Experiences”.
Por favor maxima difusión a esta perla del dictador que ahora quiere negar sus propias palabras
Claro que el dijo que ganaría las elecciones a las buenas o a las malas todos los vimos y lo escuchamos
Miren esto 👇
Rechazo vehemente el atentado contra la
vida de nuestro colega y compañero docente el Dr. Joe Cabrera. El gremio de abogados exigimos a las autoridades se esclarezcan estos hechos y que se brinden garantías para el ejercicio de nuestra profesión.
Mis condolencias a su familia.
@kb_ql22@EdmundoGU Jajajaja chavista tenía que ser, yo creo que hasta Chávez le daría vergüenza de este Nicolás !, el único fascista genocida vende patria es maduro es el mayor dictador de este siglo y aún estos están ciego!
Los informes del Panel de Expertos de la ONU y del Centro Carter, invitados internacionales del CNE, confirman la falta de transparencia en los resultados anunciados y la veracidad de las actas publicadas, que demuestran nuestro triunfo indiscutible. La voluntad del pueblo de cambiar en paz y vivir mejor, expresada el 28 de julio, es sagrada y debe ser respetada.
Venezolanos! No nos olvidemos de Freddy Superlano. El régimen lo esta torturando cruelmente para incriminarlo a él mismo y al equipo de de María Corina. Lo peor para una persona secuestrada es el olvido. Entre todos alcemos nuestra voz! #LibertadParaFreddySuperlano@freddysuperlano
Podremos contarles a nuestros nietos que vimos a Steph Curry, LeBron James y Kevin Durant jugando juntos. Podremos decirles que los vimos ganando LA MEDALLA DE ORO en los Juegos Olímpicos de París 2024. LA FOTO DEFINITIVA.