Developer Relations Engineer @Tenki, @AIsaOneHQ, @NebiusAI, Founder of @DablClub, previously @Cisco @Accenture
San Francisco, CA
Joined October 2021
- Tweets6.5K
- Following1.1K
- Followers41.1K
- Likes4.1K
Colin Lowenberg retweeted
Here's the trick, @tenki ships a @Tiny_Fish image so you can give your agent it's own backyard.
Colin Lowenberg retweeted
Tenki sandboxes can now join your @Tailscale tailnet. Your ACLs decide what an agent can reach. Ephemeral sandboxes leave your tailnet when they shut down.
Colin Lowenberg retweeted
every @tenki sandbox boots with claude code, codex and opencode already on PATH.
python 3.12, node 24, bun, gh and ripgrep too.
create a sandbox, sign in, run your agent. nothing to install.
ALT Terminal: tenki sandbox create starts a sandbox (2 CPU, 4096 MB, RUNNING). tenki sandbox exec then prints the preinstalled versions: claude 2.1.267 (Claude Code), codex-cli 0.154.0, opencode 1.17.20. Status SUCCEEDED, exit code 0.
I beat Jev at Tic-tac-toe! Thank you @typesafeai for building the model. Thank you @tenki for the sandbox.
2
506
WILD. I'm using @Tailscale to connect all my machines, and now that extends to all my agents and VMs.
New: @Tailscale connections for tenki.cloud runners and sandboxes 🔒
- Connect your tailnet once: add a Tailscale connection under Settings → Networking. It uses OpenID Connect, so Tenki stores no Tailscale auth key or OAuth secret.
-GitHub runners: make the connection the runner default for chosen repositories, or create a dedicated tenki-tailnet-* runner label. Jobs on that label join your tailnet before the first step, with no Tailscale step or GitHub secret in the workflow.
- Job access controls: limit tailnet access to protected branches, same-repository jobs, or all jobs including fork PRs.
- Sandboxes: set the connection as the default for every new sandbox, or attach it to one sandbox with tenki sandbox create --tailnet-profile-id <connection-id>. Auth keys still work if you prefer them.
- Built-in connection test: check that the connection joins your tailnet and can reach a private endpoint before you route CI through it.
📖 Docs:
- Runners and saved connections: tenki.cloud/docs/runners/tai…
- Sandboxes: tenki.cloud/docs/sandbox/tai…
@tenki is building the true enterprise cloud for agents
Colin Lowenberg retweeted
New on @tenki over the past couple of weeks:
4 new pre-built sandbox images, ready to run:
- @ampcode: a coding agent
- @PrimeIntellect's Prime Agent: a headless agent with a persistent Python kernel
- @Tiny_Fish: web agents, remote browsers, fetch and search
- @hotdata_dev: an instant SQL database for each agent task
Control what a sandbox can reach:
- Named egress policies, saved once per workspace
- HTTP rules by domain, method and path
- HTTPS inspection
- A new Egress page in the dashboard
Code Reviewer now works on @cursor_ai Origin, as well as GitHub. Connect from Tenki or from Cursor's App Marketplace, and reviews show up as inline threads with a verdict.
Preview URLs now belong to your workspace, not to one sandbox. Unbind a URL before you terminate the session, bind it to the next one, and the hostname stays the same. That's useful for OAuth callbacks and webhooks.
Also:
- Template builds show what each build uses, and you can delete builds one at a time
- Every Add Funds top-up now comes with a Stripe invoice
- Sticky disks: $0.20/GiB-mo
tenki.cloud
I fixed my CI by following @guzmanpintos all I did was copy paste this X post to Claude and my wait times dropped 50% or more
Here's the recipe to fix your CI. Give it to your agent and have it move your workflows to @tenki
ALT Table of CI job times before and after each change. TS unit tests in 4 shards: 226s to 48s. Stop running the same Go tests twice: 269s to 54s. Go integration packages in parallel: 216s to 111s. Go lint: 59s to 31s. TS lint: 61s to 40s.
1
5
779
Colin Lowenberg retweeted
Here's the recipe to fix your CI. Give it to your agent and have it move your workflows to @tenki
ALT Table of CI job times before and after each change. TS unit tests in 4 shards: 226s to 48s. Stop running the same Go tests twice: 269s to 54s. Go integration packages in parallel: 216s to 111s. Go lint: 59s to 31s. TS lint: 61s to 40s.
Colin Lowenberg retweeted
Introducing LinkedIn for your Grok @bot.
Your bot can now:
> Read your inbox and tell you who's waiting
> Draft replies from a quick note
> Welcome new connections
> Send connection requests with a personal note
You approve every message before it sends.
Pay-per-call. No subscription
Colin Lowenberg retweeted
The /goal-cup judges:
@obra @motdotla @guzmanpintos @opencolin @wschenk @p3ob7o @batuhan, Dennis Groves of OWASP, and @boardyai as AI guest judge.
They read your prompt, not only your product.
agentcommunity.org/goalcup
🤖 Made with AI
Colin Lowenberg retweeted
We’ve raised $1.3M from YC & Afore Capital so you never have to type or speak again.
Typing is slow. Talking into your computer is awkward. Subvocal fixes both
Our glasses give you the speed of speech without you making a sound or touching a keyboard.
It tracks your internal speech muscles with non-contact sensing and decodes silent, almost invisible mouthing into text at more than 200 words/min.
Pre-orders are now live at:
subvocal.co/preorder?utm_sou…
Akash changed the course of my career. I learned so much about cloud hosting, documentation, CLI tools, ephemeral sandboxes, storage and eventually GPUs
Colin Lowenberg retweeted
"Compute markets are converging." – Nick Hansen, CEO
The same miners, power sites, and capital that served Bitcoin mining are now powering AI. Here's where our $16M round led by @CMT_Digital fits in.
Colin Lowenberg retweeted
483 days ago, @5aitec and I raised $20M from @binnybansal to build ShopOS
Millions of brand images and videos later we learned: the model was never the hard part. Getting it right the 10,000th time is
Today, we’re launching @sloosh_ai to give every creator that secret sauce 👇
Colin Lowenberg retweeted
back in may when we did our first edition of agents & bagels, i ate the leftovers alone. fast forward to october our 16th edition, we ran out of bagels 😄
SF @Techweek_ edition of @scalekitinc's signature agents & bagels with @nebiusai @tavilyai @tenki 🥯
Colin Lowenberg retweeted
gm, don't blink or you'll miss it
250 sandboxes (500 vCPU and 1 TiB of RAM) provisioned in ~1s and destroyed in ~500ms
tenki.cloud is built for elastic compute. Workloads can grab burst CPU on demand, at any scale
Perfect for AI agents, RL rollouts, evals, and CI: anything that goes from idle to all cores in seconds
