Launched this week
BlickState

BlickState

AI Agents: Stop guessing why they fail. Rewind memory.

2 followers

Visit website
Stop guessing why your agents failed. Restore memory state to the exact millisecond of failure — inspect variables, check objects, and debug like you're in a Jupyter notebook at the moment of crash. Agent tool execution is a blackbox. LLMops / tracing only logs inputs and output, not state. Code runs in secure VMs — Your functions execute in isolated Firecracker microVMs with automatic checkpointing Add @tool decorator — Mark functions that should run in BlickState's secure environment
BlickState gallery image
BlickState gallery image
Launch tags:Developer Tools•Artificial Intelligence•SDK
Launch Team
agent by Firecrawl
agent by Firecrawl
Gather structured data wherever it lives on the web
Promoted

What do you think? …

Chris
Maker
📌
Thanks for visiting my page. I built BlickState due to my own frustration working with existing LLMOps solutions for tracing. LLMOps basically just logs inputs/outputs. This was fine for GenAI v1 for simple prompts / summarisations. However we are now in the next phase of Agentic AI where there are multiple tool calls and multiple Agents. Rather than spending a day digging through traces to pinpoint a tool call execution failure or a context pass failure between agents, I wanted something intuitive to what data scientists would commonly use within a Jupyter notebook or RStudio whereby the memory state (variable objects) are present and inspectable leading to faster resolution times in agent debugging. I launched BlickState (technical preview) to see if there are others who are building agents with existing LLMOps/tracing platforms are experiencing similar problems.