Hacker News: Tarcroi

New comment by Tarcroi in "Anthropic downgraded cache TTL on March 6th"

Tarcroi — Sun, 12 Apr 2026 07:39:12 +0000

This coincides with Anthropic's peak-hour announcement (March 26th). Could the throttling be partly a response to infrastructure load that was itself inflated by the TTL regression?

New comment by Tarcroi in "Show HN: Skrun – Deploy any agent skill as an API"

Tarcroi — Fri, 10 Apr 2026 10:01:00 +0000

Different thing. SkillKit distributes skills to agents. Skrun runs them as APIs.

New comment by Tarcroi in "Claude Managed Agents"

Tarcroi — Thu, 09 Apr 2026 17:44:50 +0000

As I mentioned in another comment here, I've been working on an open-source alternative. Multi-model, 5 providers with fallback. Happy to share the repo if you're interested.

New comment by Tarcroi in "Show HN: Skrun – Deploy any agent skill as an API"

Tarcroi — Thu, 09 Apr 2026 17:26:32 +0000

I agree on the commodity point, that's why I went multi-model from start.

The registry question is the one I'm thinking about the most. Right now it's flat. I plan to integrate usage data (success rates, cost, trust scores). So the registry tells you which skills actually work well, and that's valuable.

Your article looks interesting, I'll read it.

New comment by Tarcroi in "Claude Managed Agents"

Tarcroi — Wed, 08 Apr 2026 22:18:18 +0000

I've been building exactly this. it's Open source, multi-model (5 providers with fallback), from now, it runs locally but the architecture is designed for self-hosted deployment.

New comment by Tarcroi in "Show HN: Skrun – Deploy any agent skill as an API"

Tarcroi — Wed, 08 Apr 2026 21:56:23 +0000

Ha, thanks for the correction! I'll remember that!

New comment by Tarcroi in "Show HN: Skrun – Deploy any agent skill as an API"

Tarcroi — Wed, 08 Apr 2026 21:48:36 +0000

You're right. For now, it's only local. For a public deployment, the idea is to have sandboxes and verification steps. That won't completely eliminate the risk of prompt injection, but so far no solution has managed to completely resolve this problem.

New comment by Tarcroi in "Show HN: Heron is open-source security auditor that interviews your AI agents"

Tarcroi — Wed, 08 Apr 2026 21:34:53 +0000

Both. In my opinion, an agent has a life cycle and needs observability.

New comment by Tarcroi in "Show HN: Skrun – Deploy any agent skill as an API"

Tarcroi — Wed, 08 Apr 2026 21:24:55 +0000

Thanks! Sandbox deployment is planned in the roadmap. I already have a RuntimeAdapter interface in my architecture that I'll use to isolate the VMs. I'm doing exactly the same thing: I'm cross-referencing the models to challenge their plan, and my code reviewer agent's API is a big help.

New comment by Tarcroi in "Show HN: Heron is open-source security auditor that interviews your AI agents"

Tarcroi — Wed, 08 Apr 2026 13:41:51 +0000

Can this be run a second time and compared against a previous audit?

New comment by Tarcroi in "Show HN: Skrun – Deploy any agent skill as an API"

Tarcroi — Wed, 08 Apr 2026 12:50:30 +0000

Hi, I'm the "colleague", Impatient to have your feedback!