{"data":{"id":"4318efd6-a0d0-4f2e-9944-dc34693d61b2","title":"research-llm-apis 2026-04-04","summary":"A developer is redesigning the abstraction layer (a simplified interface that handles communication with many different AI services) of their LLM Python library to support new vendor features like server-side tool execution (where the AI provider runs code on their servers rather than the user's computer). They used Claude Code to analyze Python client libraries from major AI vendors and generate test commands to understand how these services handle both streaming (real-time data flow) and non-streaming data across different scenarios.","solution":"N/A -- no mitigation discussed in source.","labels":["research"],"sourceUrl":"https://simonwillison.net/2026/Apr/5/research-llm-apis/#atom-everything","publishedAt":"2026-04-05T00:32:11.000Z","cveId":null,"cweIds":null,"cvssScore":null,"cvssSeverity":null,"severity":"info","attackType":[],"issueType":"news","affectedPackages":null,"affectedVendors":["Anthropic","OpenAI","Google","Mistral"],"affectedVendorsRaw":["Anthropic","Claude","OpenAI","GPT","Google","Gemini","Mistral"],"classifierModel":"claude-haiku-4-5-20251001","classifierPromptVersion":"v3","cvssVector":null,"attackVector":null,"attackComplexity":null,"privilegesRequired":null,"userInteraction":null,"exploitMaturity":null,"epssScore":null,"patchAvailable":null,"disclosureDate":"2026-04-05T00:32:11.000Z","capecIds":null,"crossRefCount":0,"attackSophistication":"moderate","impactType":null,"aiComponentTargeted":"api","llmSpecific":true,"classifierConfidence":0.75,"researchCategory":null,"atlasIds":null}}