围绕 GPT-6 Astra 搭建 agent 操作系统的 8 个 GitHub 仓库
一条推文列出 8 个仓库，覆盖 agent 框架、记忆、工具调用、沙箱执行、评测、可观测性和上下文管道，主张单靠模型不够，需要一整套系统。
GPT 6 Astra needs a f…cking operating system around it 8 GitHub repos to build the stack around one model 01 OpenAI Agents SDK ▸ https://github.com/openai/openai-agents-python → agents + handoffs + guardrails + tracing MEMORY 02 Mem0 ▸ https://github.com/mem0ai/mem0 → persistent memory across sessions 03 Graphiti ▸ https://github.com/getzep/graphiti → temporal knowledge for changing facts TOOLS 04 Composio ▸ https://github.com/ComposioHQ/composio → 1000+ toolkits + authentication + actions EXECUTION 05 E2B ▸ https://github.com/e2b-dev/E2B → isolated environments for agent code EVALS 06 DeepEval ▸ https://github.com/confident-ai/deepeval → test complete agent trajectories OBSERVABILITY 07 Langfuse ▸ https://github.com/langfuse/langfuse → traces + evals + production debugging CONTEXT 08 LlamaIndex ▸ https://github.com/run-llama/llama_index → retrieval + context pipelines the architecture: input → context → agent → handoff → memory → knowledge → tools → sandbox → eval → trace → retry the model decides how smart the system
这条是引用着另一条写的：x.com/promixyz/status/2102424036416348494