NewsMicrosoft AI Toolkit for VS Code v0.30Microsoft Community Hub·medium signalXBlueskyLinkedInCopy linkTool Catalog, Agent Inspector, and evals-as-tests for debugging agent behavior in VS Code IDE.SourceSource pageMicrosoft Community Hub↳ Follow the threadShared entity / Stack layerMicrosoft Agent Framework .NET 1.18.0 bounds the tool-approval auto-approval loop and scopes A2A task stores by isolation keyGitHub (microsoft/agent-framework)Stack layer / Threat patternHow you lay out your repo changes prompt-injection success: highly modular workspaces measurably lower attack success ratearXiv 2608.14876Policy dependency / Stack layerCompose agent guardrails as an algebra instead of a rule list: 94.8% of policy-violating events intercepted while keeping 86.9% task completionarXiv 2608.16402Policy dependency / Stack layerKent C. Dodds ships agent-discovery endpoints in Kody: /.well-known/, markdown homepages, and RFC 8288 Link headers for unconnected agentsGitHub (kentcdodds/kody)Stack layer / ContrastCursor launches Origin, its own code hosting platform, putting repos, PRs and agents in one placeCursorStack layer / Threat patternOpenAI Agents SDK 0.22.0 redacts guardrail-blocked tool output from replayable state and rejects conflicting provider configGitHub (openai/openai-agents-python)Policy dependency / Stack layerClaude Agent SDK for Python 0.2.140 adds MCP 2.x in-process servers and a structured ResultError instead of exit code 1GitHub (anthropics/claude-agent-sdk-python)Stack layer / Threat patternAmjad Masad on Replit's New Black-Box Pen Testing: 'It's Not Enough to Scan Your Code for Vulnerabilities'X/Twitter