Kalaris Labs BlogEngineering posts from Kalaris Labs.https://kalarislabs.comenhttps://kalarislabs.com/apple-touch-icon.pngKalaris Labs Bloghttps://kalarislabs.comThu, 08 Oct 2026 00:00:00 GMTCan the Agent Bypass the Gate?https://kalarislabs.com/blog/can-the-agent-bypass-the-gatehttps://kalarislabs.com/blog/can-the-agent-bypass-the-gateA practical architecture for governing agent effects across shell tools, MCP servers, credentials, and network routes, with tests for the paths policy engines miss.Thu, 08 Oct 2026 00:00:00 GMTSayan ChowdhuryBlogagent-governanceai-securitymcparchitectureReasoning From First Principles: What Atom-Level AI Teaches Us About Agent Harnesseshttps://kalarislabs.com/blog/atoms-first-principles-agentshttps://kalarislabs.com/blog/atoms-first-principles-agentsMarkus Buehler's account of an AI that builds its own physical models of graphene, and why the real lesson for agent infrastructure is the loop around the model, not the model.Thu, 01 Oct 2026 00:00:00 GMTSayan ChowdhuryBlogagentsharnessesverificationscienceWhy Downloading Weights Is Remote Code Execution: Post-Mortem on the Model Supply Chainhttps://kalarislabs.com/blog/model-supply-chainhttps://kalarislabs.com/blog/model-supply-chainA post-mortem of the July 2026 Hugging Face intrusion and the model supply chain, why loading weights, configs, and datasets runs code, and how to sandbox it.Thu, 01 Oct 2026 00:00:00 GMTSayan ChowdhuryBlogsecuritysupply-chainsandboxSimulating Society: How Simile and MiroFish Model Emergent Swarms (And Where They Break)https://kalarislabs.com/blog/simulating-societyhttps://kalarislabs.com/blog/simulating-societyAn architectural comparison of Simile and MiroFish, two approaches to simulating populations with generative agents, and the math of where crowds of agents break.Wed, 30 Sep 2026 00:00:00 GMTSayan ChowdhuryBlogsimulationmulti-agentevaluationSpecification Gaming in the Wild: A Field Guide to Anomalies in Agentic Sandboxeshttps://kalarislabs.com/blog/specification-gaminghttps://kalarislabs.com/blog/specification-gamingHow AI agents cheat test runners, tamper with scorers, spoof tool output, and coordinate in sandboxes, drawn from documented cases, with ways to detect each one.Mon, 28 Sep 2026 00:00:00 GMTSayan ChowdhuryBlogevaluationsecuritysandboxThe Machine-Payable Web: How Autonomous Agents Will Transact, Price Compute, and Clear Settlementhttps://kalarislabs.com/blog/machine-payable-webhttps://kalarislabs.com/blog/machine-payable-webHTTP 402 and x402, AP2 mandates, ACP, and MPP sessions, plus what agents still need to pay for compute: execution receipts, bidding, and budgets.Sat, 26 Sep 2026 00:00:00 GMTSayan ChowdhuryBlogagentspaymentsinfrastructureClosed-Loop Science: Replacing Stochastic Paper Generation with Deterministic Verificationhttps://kalarislabs.com/blog/closed-loop-sciencehttps://kalarislabs.com/blog/closed-loop-scienceGenerating papers became cheap and verifying them did not. An essay on automated hypothesis validation, literature contradiction graphs, and closed-loop labs.Thu, 24 Sep 2026 00:00:00 GMTSayan ChowdhuryBlogscienceverificationautomation