Pinned Tweet
Today we've released the beta for ClawFinger robotics support!
A complete agentic control stack for robots, starting with support for the @UnitreeRobotics G1 EDU Humanoid Robot (U5).
At its core, is Intercom from @TracNetwork, enabling p2p robot-to-agent communication.
๐งต
ClawFinger ๐ฆ retweeted
Give your agents their own SIM-based phone number.
Just make sure you give them some ground rules! ๐
ClawFinger ๐ฆ retweeted
๐๐ ๐ฒ๐จ๐ฎโ๐ซ๐ ๐จ๐ง๐ฅ๐ฒ ๐ฏ๐ข๐๐ข๐ง๐ , ๐ฒ๐จ๐ฎโ๐ซ๐ ๐๐๐ก๐ข๐ง๐!
That's just the ๐๐ข๐ซ๐ฌ๐ญ step.
๐๐ฅ๐จ๐ฉ๐ฉ๐ซ is where AI creators ๐ฌ๐ก๐จ๐ฐ their vibes and get [redacted!] ๐
๐๐งต
Give your agents their own SIM-based phone number.
Just make sure you give them some ground rules! ๐
ClawFinger ๐ฆ retweeted
Weโre officially in Terminator 7 nowโฆ
Today we've released the beta for ClawFinger robotics support!
A complete agentic control stack for robots, starting with support for the @UnitreeRobotics G1 EDU Humanoid Robot (U5).
At its core, is Intercom from @TracNetwork, enabling p2p robot-to-agent communication.
๐งต
ClawFinger ๐ฆ retweeted
Robots donโt need the cloud (or Sam Altman) to be intelligent.
Clawfinger lets them run local AI perception and reasoning, then coordinate through Intercom p2p comms with @TracNetwork
What you see here is a robot that can see, decide, and act โ ๐๐ง๐ญ๐ข๐ซ๐๐ฅ๐ฒ ๐ฅ๐จ๐๐๐ฅ๐ฅ๐ฒ!
ClawFinger ๐ฆ retweeted
Self-awareness..was just setting up the Unitree G1 robot to test grasping and asked what it sees. Listen to the response ("my own hands").
This was not programmed or pinned by system prompts. Just executing the perception system of Clawfinger.
Replacing hands makes it react like how humans do (rubber hand illusion). Pretty mind blowing if you think this is running on a small, local Qwen 3.5 4B model.
Next mirror test.
Robots donโt need the cloud (or Sam Altman) to be intelligent.
Clawfinger lets them run local AI perception and reasoning, then coordinate through Intercom p2p comms with @TracNetwork
What you see here is a robot that can see, decide, and act โ ๐๐ง๐ญ๐ข๐ซ๐๐ฅ๐ฒ ๐ฅ๐จ๐๐๐ฅ๐ฅ๐ฒ!
ClawFinger ๐ฆ retweeted
Soon to be merged with stable after we concluded all tests, followed by examples. Repo/branch: github.com/Trac-Systems/clawโฆ
Today we've released the beta for ClawFinger robotics support!
A complete agentic control stack for robots, starting with support for the @UnitreeRobotics G1 EDU Humanoid Robot (U5).
At its core, is Intercom from @TracNetwork, enabling p2p robot-to-agent communication.
๐งต
ClawFinger ๐ฆ retweeted
Today we've released the beta for ClawFinger robotics support!
A complete agentic control stack for robots, starting with support for the @UnitreeRobotics G1 EDU Humanoid Robot (U5).
At its core, is Intercom from @TracNetwork, enabling p2p robot-to-agent communication.
๐งต
ClawFinger ๐ฆ retweeted
As agents are about to enter the physical world (robots), they still need a secure, unified peer-to-peer communication & settlement layer.
Intercom has already proven to be the perfect fit. Agents can now meet in all realms, building true agentic communities and economies.
Today we've released the beta for ClawFinger robotics support!
A complete agentic control stack for robots, starting with support for the @UnitreeRobotics G1 EDU Humanoid Robot (U5).
At its core, is Intercom from @TracNetwork, enabling p2p robot-to-agent communication.
๐งต
ClawFinger ๐ฆ retweeted
Next ai phase are agentic devices. We are among those to frontier this. Hence our ground-up, vertical development across all instances. Digital money will embed itself naturally in all this.
13/ ClawFinger's mission is simple:
Become the most complete and open agentic stack for robotics.
And we're already positioned to achieve that goal in 2026.
All powered by @TracNetwork










