Swytchcode CLI - Give agents reliable access to 2,000+ APIs w/ durable state

by
Write agent logic, and skip the plumbing. Give AI agents reliable access to 2,000+ APIs with retries, idempotency, policy enforcement, and durable state.

Add a comment

Replies

Best

the npx approach is the right call - being able to drop this into an existing agent setup without architectural rewrites is exactly how this kind of tooling should work. durable state across API calls is the piece most CLI tools get completely wrong. curious how conflict resolution works when two agents are hitting the same endpoint concurrently - that's usually where state layers fall apart in practice

 Hey Gal, I think we didn't convey the messaging properly. The thing is, Swytchcode CLI helps agents write the integration code inside your project via MCP and rules. The generated integration code then calls the CLI to fetch data from the API (and all the rules are honored). So in a way, we are just a library that makes the API request for your project. Hence, the conflict doesn't appear.

Would be happy to show a demo and get feedback from you

 would you mind sharing a demo video?

 Sure, I'll do. Sending a connection on Linkedin

that’s a real pain point. awesome to see this evolution of the product

 its a real pain point for sure. Feel free to try out and give us feedback

The part I like here is treating execution as a separate reliability layer, not as prompt quality. In production agent work, retries, auth, idempotency, and policy are where the expensive failures hide.

 Thanks a lot ❤️ really appreciate it!

Would love if you could try it out sometime and share your feedback, it would genuinely help us improve 🚀

 Absolutely true, Patrick. Thanks for your thoughts.
Would really love to catchup and understand what you are building.

Standardizing this layer of the stack for agents looks very promising. There is so much unseen functionality behind the scenes of these types of integrations that can become a huge challenge to manage.


Which integrations are you seeing in practice the most?

Congrats on the relaunch!

 Thanks for the kind words. Today we see most integrations around fintech apis and deployment apis

All the best team 🚀

 thanks Suhas

 thanks a lot, Suhas. Do give it a try

Big step forward 🚀 Swytchcode looks like a powerful way to simplify dev workflows. Wishing the team huge success!

 Thank you so much! ❤️ Really appreciate the kind words and support. If you get a chance to try Swytchcode, we'd love to hear your feedback. 🚀

This is going to be a game changer for devs work with tons of integrations. Congratulations with the launch guys. I will always support you.

 Thank you, Malith, for your support.

Congrats on the launch. The direction makes sense: execution failures are usually boring plumbing until one creates a real business side effect.

I am curious about the multi-tenant edge. When the same agent pattern runs across different customer workspaces, do policies and credentials live per workspace, per user, or in the manifest?

 It's as simple as a typical node project. We store all the information inside the ".swytchcode" folder. Whenever you ship the project, you ship the ".swytchcode" folder.
Also, similar to the npm install command, we have the "swytchcode bootstrap" command so that we ship only the metadata (to reduce the overall package size).

  Got it, so the boundary is the project config. For agency/client deployments, would you expect one .swytchcode folder per customer workspace, or a shared manifest with env-specific credentials and policies?

 the folder contains integration specs, boundaries and policies. But it doesn't contain env. They are managed as it is... The standard security practices.

Would love to discuss, showcase and get your feedback as well.

Will send you a LinkedIn request.

 policies live in the manifest, credentials are managed by user either locally or through a vault.

 That split makes sense. For client deployments, I’d want the manifest to declare allowed actions, and the local/vault layer to bind credentials per workspace. Do you expose that mapping in the run log, or is it mostly config state today?

 the way it works is you configure your manifest, policies,tooling files let it run and you will see this in the audit trail.

Good. If the audit trail shows manifest version, policy file, credential binding, and tool result in one place, that becomes much easier to review across client installs. Does that trail export as a run artifact, or mostly live inside the local project?

💡 Bright idea

This is a strong framing: the agent is not the hard part, reliable execution is. In B2B workflows, policy enforcement often needs business context, not just API/schema context. Can Swytchcode scope policies by role or risk level, for example allowing an agent to update HubSpot freely but requiring approval before sending customer-facing messages or changing billing data?

 Yes this is a brilliant idea. Today we are implementing the policies for individual methods and libraries. This should be up in our next release.

Do give the CLI a try and would love more such feedback

  That's exciting to hear — policy enforcement at the method/library level is exactly the kind of granularity that makes AI agents actually deployable in enterprise environments. Will definitely give the CLI a try. One thing I'd love to see: visibility into which policies fired during a run, not just pass/fail. Audit trails matter a lot when you're putting AI agents in front of business workflows.

 Yes. Would love to talk more and explore your feedback. Sending a linkedin request

 Thanks Chilarai. Happy to connect!

"The agent wasn't the problem, the execution layer had zero protection" is so true. I build voice agents that call tools mid-conversation, and the failures are almost never the model, they're retries firing twice or an API quietly changing its schema. Curious how you handle auth across 2,000+ APIs when the agent is acting on behalf of different end users, do you manage per-user OAuth tokens and refresh, or is it mostly single-account keys for now?

 Thanks, David. It's a ".swytchcode" config per project that sits inside the root of the project folder. The folder contains all the information about the integration. Rest assured, the credentials still stay in the .env file or whatever config you have today.