[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-799":3},{"id":4,"name":5,"fullName":6,"owner":5,"repo":5,"description":7,"homepage":8,"htmlUrl":9,"language":10,"languages":9,"totalLinesOfCode":9,"stars":11,"forks":12,"watchers":13,"openIssues":14,"contributorsCount":15,"subscribersCount":15,"size":15,"stars1d":15,"stars7d":16,"stars30d":17,"stars90d":15,"forks30d":15,"starsTrendScore":15,"compositeScore":18,"rankGlobal":9,"rankLanguage":9,"license":19,"archived":20,"fork":20,"defaultBranch":21,"hasWiki":22,"hasPages":20,"topics":23,"createdAt":9,"pushedAt":9,"updatedAt":30,"readmeContent":31,"aiSummary":32,"trendingCount":15,"starSnapshotCount":15,"syncStatus":33,"lastSyncTime":34,"discoverSource":35},799,"future-agi","future-agi\u002Ffuture-agi","Open-source, end-to-end platform for evaluating, observing, and improving LLM and AI agent applications. Tracing · Evals · Simulations · Datasets · Gateway · Guardrails. Self-hostable. Apache 2.0.","https:\u002F\u002Ffutureagi.com",null,"Python",1126,239,9,59,0,50,216,85.14,"Apache License 2.0",false,"main",true,[24,25,26,27,28,29],"ai","ai-gateway","evals","llm","observability","simulation","2026-06-11 04:00:33","\u003C!--\n╔═════════════════════════════════════════════════════════════════════════════╗\n║  MARKETING NOTES FOR IMAGE ASSETS                                           ║\n║                                                                             ║\n║  All images below live under .github\u002Fassets\u002F. Specs + intent are inlined   ║\n║  above each \u003Cimg> tag as HTML comments. Total asset budget \u003C 12 MB.         ║\n║  Use PNG for static screenshots, GIF only where called out. Ship light +    ║\n║  dark variants via \u003Cpicture> for any image that contains a UI screenshot    ║\n║  (GitHub dark-mode users will see the dark file).                           ║\n╚═════════════════════════════════════════════════════════════════════════════╝\n-->\n\n> ⚠️ **Nightly release for early testing.** Expect rough edges. Stable version coming out soon — please open an issue if you hit anything.\n\n\u003Cdiv align=\"center\">\n\n\u003C!--\n  [MARKETING] logo-banner.png \u002F logo-banner-dark.png\n  What:    Full wordmark \"Future AGI\" + single-line tagline \"AI Agents\n           hallucinate. Fix it faster.\" — centered, brand colors.\n  Size:    1600 × 400, PNG, transparent background.\n  Variants: light + dark; swap via \u003Cpicture>.\n-->\n\u003Ca href=\"https:\u002F\u002Ffutureagi.com\">\n  \u003Cimg alt=\"Future AGI — make AI agents reliable\" src=\"frontend\u002Fpublic\u002Fassets\u002Freadme\u002FLogo.png\" width=\"100%\">\n\u003C\u002Fa>\n\n# AI Agents hallucinate. Fix it faster.\n\n**The open-source platform for shipping self-improving AI agents.** Evaluations, tracing, simulations, guardrails, gateway, optimization. Everything runs on one platform and one feedback loop, from first prototype to live deployment.\n\n\u003Cp>\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Ffuture-agi\u002Ffuture-agi\u002Fblob\u002Fmain\u002FLICENSE\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Flicense-Apache%202.0-blue?style=flat-square\" alt=\"Apache 2.0 License\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fpypi.org\u002Fproject\u002Fai-evaluation\u002F\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fpypi\u002Fv\u002Fai-evaluation?style=flat-square&label=pypi\" alt=\"PyPI\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fwww.npmjs.com\u002Fpackage\u002F@traceai\u002Ffi-core\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fnpm\u002Fv\u002F@traceai\u002Ffi-core?style=flat-square&label=npm\" alt=\"npm\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fdiscord.gg\u002FUjZ2gRT5p\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fdiscord-join-5865F2?style=flat-square\" alt=\"Discord\">\u003C\u002Fa>\n\u003C\u002Fp>\n\n\u003Cp>\n  \u003Ca href=\"https:\u002F\u002Fapp.futureagi.com\u002Fauth\u002Fjwt\u002Fregister\">\u003Cb>Try Cloud (Free)\u003C\u002Fb>\u003C\u002Fa> ·\n  \u003Ca href=\"#-quickstart-60-seconds\">\u003Cb>Self-Host\u003C\u002Fb>\u003C\u002Fa> ·\n  \u003Ca href=\"https:\u002F\u002Fdocs.futureagi.com\">\u003Cb>Docs\u003C\u002Fb>\u003C\u002Fa> ·\n  \u003Ca href=\"https:\u002F\u002Ffutureagi.com\u002Fblog\">\u003Cb>Blog\u003C\u002Fb>\u003C\u002Fa> ·\n  \u003Ca href=\"https:\u002F\u002Fdiscord.gg\u002FUjZ2gRT5p\">\u003Cb>Discord\u003C\u002Fb>\u003C\u002Fa> ·\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Forgs\u002Ffuture-agi\u002Fdiscussions\">\u003Cb>Discussions\u003C\u002Fb>\u003C\u002Fa>\n\u003C\u002Fp>\n\n\u003C\u002Fdiv>\n\n---\n\n\u003C!--\n  [MARKETING] hero-demo.gif\n  What:    8–12 second product loop. Suggested flow:\n             1. Open a trace → expand a span\n             2. Click \"Run Eval\" → score appears\n             3. Navigate to Simulate → a voice agent conversation plays\n             4. Navigate to Gateway → cost + guardrail dashboard\n  Size:    1600 × 900, GIF ≤ 4 MB (use `gifski` to compress).\n  Fallback: hero-demo.png (first frame) for users who disable GIFs.\n  Why here: carries the \"what is this?\" answer in 3 seconds; biggest\n            single lift on time-on-page across flagship OSS READMEs.\n-->\n\u003Cdiv align=\"center\">\n  \u003Cimg alt=\"Future AGI — trace an agent, run evals, simulate, and guardrail in one platform\" src=\"frontend\u002Fpublic\u002Fassets\u002Freadme\u002Fself-host.gif\" width=\"720\">\n\u003C\u002Fdiv>\n\n---\n\n## Why Future AGI?\n\nMost AI agents fail in production, and teams end up stitching together evals, observability, and guardrails that never close the loop.\nFutureAGI collapses all of it into one platform and one feedback loop. Simulate edge cases before launch, evaluate what happens in production, protect users in real time, and turn every trace into signal for the next version.\nThe result: agents that don't just get monitored, they self-improve.\n\n\u003Ctable>\n\u003Ctr>\n\u003Ctd width=\"33%\" valign=\"top\">\n\n###  All-in-one\nNo more stitching Langfuse + Braintrust + Helicone + Guardrails AI + a custom simulator. One platform covers the lifecycle: **simulate → evaluate → protect → monitor → optimize**, with data flowing back as a loop.\n\n\u003C\u002Ftd>\n\u003Ctd width=\"33%\" valign=\"top\">\n\n###  Open & self-hostable\nApache 2.0 core. Every evaluator, every prompt, every trace is inspectable — **no black-box scoring**. Self-host for data sovereignty or use our managed Cloud. Drop in your own stack at any layer via OTel \u002F OpenAI-compatible HTTP.\n\n\u003C\u002Ftd>\n\u003Ctd width=\"33%\" valign=\"top\">\n\n###  Built for production\nGo-based gateway with **~9.9 ns weighted routing**, **~29 k req\u002Fs on t3.xlarge**, **P99 ≤ 21 ms with guardrails on**. OpenTelemetry-native traces. 50+ framework instrumentors. Every claim reproducible via the committed benchmark harness.\n\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftable>\n\n---\n\n## 🚀 Quickstart (60 seconds)\n\nThree ways, picked by how much you want to install:\n\n\u003Ctable width=\"100%\">\n\u003Ctr>\n\u003Cth width=\"34%\">Cloud (fastest)\u003C\u002Fth>\n\u003Cth width=\"36%\">Self-host (Docker)\u003C\u002Fth>\n\u003Cth width=\"30%\">Self-host (Kubernetes)\u003C\u002Fth>\n\u003C\u002Ftr>\n\u003Ctr valign=\"top\">\n\u003Ctd width=\"34%\">\n\n**No install. Free tier.**\n\n```bash\n# Sign up free:\n#   app.futureagi.com\n\npip install ai-evaluation\n```\n\n\u003Csub>SOC 2 Type II · HIPAA · data stays in your region.\u003C\u002Fsub>\n\n\u003C\u002Ftd>\n\u003Ctd width=\"36%\">\n\n**One command, full stack.**\n\n```bash\ngit clone https:\u002F\u002Fgithub.com\u002Ffuture-agi\u002Ffuture-agi.git\ncd future-agi\ncp futureagi\u002F.env.example futureagi\u002F.env\ndocker compose up -d\n```\n\nOpen [http:\u002F\u002Flocalhost:3031](http:\u002F\u002Flocalhost:3031).\n\n\u003C\u002Ftd>\n\u003Ctd width=\"30%\">\n\n**Production-grade, HA.**\n\n```bash\nhelm repo add futureagi \\\nhelm install fagi futureagi\u002Ffuture-agi\n```\n\n\u003Csub>Helm chart — v1 in progress. Until then, kubectl manifests in `deploy\u002F`.\u003C\u002Fsub>\n\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftable>\n\n### Instrument your first agent\n\n\u003Ctable width=\"100%\">\n\u003Ctr>\n\u003Ctd width=\"50%\">\n\n**Python**\n```python\nfrom fi_instrumentation import register\nfrom traceai_openai import OpenAIInstrumentor\n\nregister(project_name=\"my-agent\")\nOpenAIInstrumentor().instrument()\n\n# Your existing OpenAI code is now traced.\nresponse = client.chat.completions.create(\n    model=\"gpt-4o\",\n    messages=[{\"role\": \"user\", \"content\": query}],\n)\n```\n\n\u003C\u002Ftd>\n\u003Ctd width=\"50%\">\n\n**TypeScript**\n```typescript\nimport { register } from \"@traceai\u002Ffi-core\";\nimport { OpenAIInstrumentation } from \"@traceai\u002Fopenai\";\n\nregister({ projectName: \"my-agent\" });\nnew OpenAIInstrumentation().instrument();\n\n\u002F\u002F Your existing OpenAI code is now traced.\nconst response = await openai.chat.completions.create({\n  model: \"gpt-4o\",\n  messages: [{ role: \"user\", content: query }],\n});\n```\n\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftable>\n\n\u003Csub> [Full docs →](https:\u002F\u002Fdocs.futureagi.com)  ·  [Cookbooks →](https:\u002F\u002Fdocs.futureagi.com\u002Fdocs\u002Fcookbook)  ·  [API reference →](https:\u002F\u002Fdocs.futureagi.com\u002Fdocs\u002Fapi)\u003C\u002Fsub>\n\n---\n\n## Core features\n\nSix pillars. Each one replaces a tool you probably have.\n\n\u003Ctable>\n\u003Ctr>\n\u003Ctd width=\"33%\" valign=\"top\">\n\n### 🧪 Simulate\nThousands of multi-turn conversations against realistic personas, adversarial inputs, and edge cases. Text **and voice** (LiveKit, VAPI, Retell, Pipecat).\n\n\u003Csub>[Docs →](https:\u002F\u002Fdocs.futureagi.com\u002Fdocs\u002Fsimulation)\u003C\u002Fsub>\n\n\u003C\u002Ftd>\n\u003Ctd width=\"33%\" valign=\"top\">\n\n### 📊 Evaluate\n50+ metrics under one `evaluate()` call: groundedness, hallucination, tool-use correctness, PII, tone, custom rubrics. **LLM-as-judge + heuristic + ML.**\n\n\u003Csub>[Docs →](https:\u002F\u002Fdocs.futureagi.com\u002Fdocs\u002Fevaluation)\u003C\u002Fsub>\n\n\u003C\u002Ftd>\n\u003Ctd width=\"33%\" valign=\"top\">\n\n### 🛡️ Protect\n18 built-in scanners (PII, jailbreak, injection, …) + 15 vendor adapters (Lakera, Presidio, Llama Guard, …). Inline in gateway or standalone SDK.\n\n\u003Csub>[Docs →](https:\u002F\u002Fdocs.futureagi.com\u002Fdocs\u002Fprotect)\u003C\u002Fsub>\n\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd width=\"33%\" valign=\"top\">\n\n### 👁️ Monitor\nOpenTelemetry-native tracing across 50+ frameworks (LangChain, LlamaIndex, CrewAI, DSPy…). Span graphs, latency, token cost, live dashboards. Zero-config.\n\n\u003Csub>[Docs →](https:\u002F\u002Fdocs.futureagi.com\u002Fdocs\u002Fobserve)\u003C\u002Fsub>\n\n\u003C\u002Ftd>\n\u003Ctd width=\"33%\" valign=\"top\">\n\n### 🎛️ Agent Command Center\nOpenAI-compatible gateway. 100+ providers, 15 routing strategies, semantic caching, virtual keys, MCP, A2A. **~29k req\u002Fs, P99 ≤ 21ms with guardrails on.**\n\n\u003Csub>[Docs →](https:\u002F\u002Fdocs.futureagi.com\u002Fdocs\u002Fcommand-center) · [Benchmarks →](.\u002Ffutureagi\u002Fagentcc-gateway\u002FREADME.md#-benchmarks)\u003C\u002Fsub>\n\n\u003C\u002Ftd>\n\u003Ctd width=\"33%\" valign=\"top\">\n\n### 🔁 Optimize\nSix prompt-optimization algorithms (GEPA, PromptWizard, ProTeGi, Bayesian, Meta-Prompt, Random). Production traces feed back as training data.\n\n\u003Csub>[Docs →](https:\u002F\u002Fdocs.futureagi.com\u002Fdocs\u002Foptimization)\u003C\u002Fsub>\n\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftable>\n\n---\n\n##  Deployment options\n\n\u003C!--\n  [MARKETING] deploy-buttons.png  (optional — can stay as inline shields)\n  What:    A horizontal row of one-click-deploy badges: Docker · Render ·\n           Railway · Fly · AWS Marketplace (coming soon).\n  Size:    1400 × 120, PNG — OR keep as inline \u003Ca>\u003Cimg>\u003C\u002Fa> badges below.\n-->\n\u003C!--\n\u003Cp align=\"center\">\n  \u003Ca href=\"https:\u002F\u002Frender.com\u002Fdeploy\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fdeploy%20on-Render-46E3B7?style=for-the-badge&logo=render&logoColor=white\" alt=\"Deploy on Render\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Ffly.io\u002Fdocs\u002Flaunch\u002F\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fdeploy%20on-Fly-7A3CEF?style=for-the-badge&logo=flydotio&logoColor=white\" alt=\"Deploy on Fly\">\u003C\u002Fa>\n  \u003Ca href=\"#-quickstart-60-seconds\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fdocker%20compose-up%20-d-2496ED?style=for-the-badge&logo=docker&logoColor=white\" alt=\"Docker Compose\">\u003C\u002Fa>\n\u003C\u002Fp>\n-->\n| Target | Status | Notes |\n|---|:---:|---|\n|  Docker Compose | ✅ | `docker compose up -d` from a fresh clone |\n|  Kubernetes | ✅ | Plain manifests today; Helm chart v1 in progress |\n|  AWS \u002F GCP \u002F Azure | ✅ | Runs on any container runtime — ECS · Cloud Run · AKS · EKS · GKE |\n|  AWS Marketplace | ⏳ | Coming soon |\n|  Air-gapped \u002F on-prem | ✅ | No phone-home — [contact sales](mailto:sales@futureagi.com) |\n\n---\n\n##  Architecture\n\nEvery arrow is an open, documented interface: **OpenTelemetry OTLP** for traces, **OpenAI-compatible HTTP** for the gateway, **Postgres \u002F ClickHouse SQL** for storage. Drop in your own stack at any layer.\n\n\u003C!--\n  [MARKETING] architecture.svg  (already exists — leave as-is unless re-designing)\n  What:    4-band system diagram: client layer → edge (traceAI + gateway)\n           → platform (simulate · eval · monitor · optimize) → data layer.\n  Size:    ~1400w, vector SVG (existing file 1200×760 in dark palette).\n-->\n\u003C!-- \u003Cpicture>\n  \u003Csource media=\"(prefers-color-scheme: dark)\" srcset=\".github\u002Fassets\u002Farchitecture.svg\">\n  \u003Cimg alt=\"Future AGI architecture — client SDKs → traceAI + Agent Command Center → Django platform → PostgreSQL, ClickHouse, Redis, RabbitMQ\" src=\".github\u002Fassets\u002Farchitecture.svg\" width=\"100%\">\n\u003C\u002Fpicture> -->\n\n**Runtime:** Python 3.11+ (Django 4.2 + Channels) · Go 1.23+ (gateway) · React 18 + Vite · Node 20+.\n**Data:** PostgreSQL (metadata) · ClickHouse (spans + time-series) · Redis (state) · RabbitMQ + Temporal (jobs).\n\n\u003Cdetails>\u003Csummary>Component breakdown (per-package)\u003C\u002Fsummary>\n\n| Layer | Component | Code |\n|---|---|---|\n|  Edge | **traceAI** — OpenTelemetry instrumentation | [`future-agi\u002FtraceAI`](https:\u002F\u002Fgithub.com\u002Ffuture-agi\u002FtraceAI) |\n|  Edge | **Agent Command Center** — OpenAI-compatible proxy | [`futureagi\u002Fagentcc-gateway\u002F`](.\u002Ffutureagi\u002Fagentcc-gateway) |\n|  Platform | **tracer** — OTLP ingest, span graph | [`futureagi\u002Ftracer\u002F`](.\u002Ffutureagi\u002Ftracer) |\n|  Platform | **agentic_eval** — 50+ metrics, LLM-as-judge | [`futureagi\u002Fagentic_eval\u002F`](.\u002Ffutureagi\u002Fagentic_eval) |\n|  Platform | **simulate** — persona-driven scenario generation | [`futureagi\u002Fsimulate\u002F`](.\u002Ffutureagi\u002Fsimulate) |\n|  Platform | **model_hub** — LLM routing, embeddings, datasets | [`futureagi\u002Fmodel_hub\u002F`](.\u002Ffutureagi\u002Fmodel_hub) |\n|  Platform | **accounts · usage · integrations** — auth, orgs, metering, connectors | [`futureagi\u002Faccounts\u002F`](.\u002Ffutureagi\u002Faccounts) |\n|  Data | **PostgreSQL** · **ClickHouse** · **Redis** · **RabbitMQ + Temporal** | — |\n\n\u003C\u002Fdetails>\n\n---\n\n##  SDKs & integrations\n\nFuture AGI is an **open-source ecosystem** — each SDK is independently usable, independently packaged, Apache\u002FMIT-licensed.\n\n### Client libraries\n\n| Repo | Install | Languages | Purpose |\n|---|---|---|---|\n| [**traceAI**](https:\u002F\u002Fgithub.com\u002Ffuture-agi\u002FtraceAI) | `pip install fi-instrumentation-otel`\u003Cbr>`npm i @traceai\u002Ffi-core` | Python · TS · Java · C# | **Zero-config OTel tracing** for 50+ AI frameworks |\n| [**ai-evaluation**](https:\u002F\u002Fgithub.com\u002Ffuture-agi\u002Fai-evaluation) | `pip install ai-evaluation`\u003Cbr>`npm i @future-agi\u002Fai-evaluation` | Python · TS | **50+ evaluation metrics** + guardrail scanners |\n| [**futureagi**](https:\u002F\u002Fgithub.com\u002Ffuture-agi\u002Ffutureagi-sdk) | `pip install futureagi` | Python | Platform SDK — datasets, prompts, KB, experiments |\n| [**agent-opt**](https:\u002F\u002Fgithub.com\u002Ffuture-agi\u002Fagent-opt) | `pip install agent-opt` | Python | **6 prompt-optimization algorithms** (GEPA, PromptWizard, …) |\n| [**simulate-sdk**](https:\u002F\u002Fgithub.com\u002Ffuture-agi\u002Fsimulate-sdk) | `pip install agent-simulate` | Python | Voice-agent simulation via LiveKit + Silero VAD |\n| [**agentcc**](https:\u002F\u002Fgithub.com\u002Ffuture-agi\u002Fagent-command-center-sdk) | `pip install agentcc`\u003Cbr>`npm i @agentcc\u002Fclient` | Python · TS (+ LangChain · LlamaIndex · React · Vercel) | Gateway client SDKs |\n\n### Integrations\n\n\u003C!--\n  [MARKETING] integrations-grid.png\n  What:    5×4 logo grid, grayscale. Rows roughly: LLM providers,\n           frameworks, voice platforms, vector DBs, tools.\n           Contents (keep grayscale — colored logos look like an ad):\n             LLM:       OpenAI · Anthropic · Google · AWS Bedrock · Azure\n             Framework: LangChain · LlamaIndex · CrewAI · AutoGen · DSPy\n             Voice:     LiveKit · VAPI · Retell · Pipecat · Deepgram\n             Vector:    Pinecone · Qdrant · Weaviate · Chroma · Milvus\n             Tools:     OpenTelemetry · Vercel · MCP · A2A · HuggingFace\n  Size:    1600 × 800, PNG — OR keep the current 6-row markdown table\n           below as fallback.\n-->\n\u003C!--\n\u003Cdiv align=\"center\">\n  \u003Cimg alt=\"Future AGI integrations — 100+ LLM providers, 50+ frameworks, every major vector DB and voice platform\" src=\".github\u002Fassets\u002Fintegrations-grid.png\" width=\"100%\">\n\u003C\u002Fdiv>\n-->\n| | |\n|---|---|\n| **LLM providers (100+)** | OpenAI · Anthropic · Google Gemini · Vertex AI · AWS Bedrock · Azure OpenAI · Mistral · Groq · Cohere · Together · Perplexity · OpenRouter · Fireworks · xAI · Replicate · HuggingFace · + self-hosted **Ollama · vLLM · LM Studio · TGI · Llamafile** |\n| **Agent frameworks** | LangChain · LangGraph · LlamaIndex · CrewAI · AutoGen · Phidata · PydanticAI · Claude SDK · LiteLLM · Haystack · DSPy · Instructor · Smol-agents |\n| **Voice platforms** | VAPI · Retell · LiveKit · Pipecat |\n| **Vector DBs** | Pinecone · Weaviate · Chroma · Milvus · Qdrant · pgvector |\n| **Tools & infra** | Vercel AI SDK · n8n · MongoDB · MCP · A2A · Guardrails AI · Langfuse · HuggingFace Smol-agents |\n\n\u003Csub> [Full integrations catalog →](https:\u002F\u002Fdocs.futureagi.com\u002Fdocs\u002Fintegrations)\u003C\u002Fsub>\n\n---\n\n##  How Future AGI compares\n\n\u003Ctable width=\"100%\">\n\u003Cthead>\n\u003Ctr>\n\u003Cth width=\"25%\">\u003C\u002Fth>\n\u003Cth width=\"15%\" align=\"center\">\u003Cb>Future AGI\u003C\u002Fb>\u003C\u002Fth>\n\u003Cth width=\"15%\" align=\"center\">Langfuse\u003C\u002Fth>\n\u003Cth width=\"15%\" align=\"center\">Phoenix\u003C\u002Fth>\n\u003Cth width=\"15%\" align=\"center\">Braintrust\u003C\u002Fth>\n\u003Cth width=\"15%\" align=\"center\">Helicone\u003C\u002Fth>\n\u003C\u002Ftr>\n\u003C\u002Fthead>\n\u003Ctbody>\n\u003Ctr>\u003Ctd>Open source\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003Cbr>\u003Csub>Apache 2.0\u003C\u002Fsub>\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003Cbr>\u003Csub>MIT\u003C\u002Fsub>\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003Cbr>\u003Csub>Elastic v2\u003C\u002Fsub>\u003C\u002Ftd>\u003Ctd align=\"center\">❌\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003Cbr>\u003Csub>Apache 2.0\u003C\u002Fsub>\u003C\u002Ftd>\u003C\u002Ftr>\n\u003Ctr>\u003Ctd>Self-host\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003Ctd align=\"center\">❌\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003C\u002Ftr>\n\u003Ctr>\u003Ctd>LLM tracing (OpenTelemetry)\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003Ctd align=\"center\">⚠️\u003Cbr>\u003Csub>via OpenLLMetry\u003C\u002Fsub>\u003C\u002Ftd>\u003C\u002Ftr>\n\u003Ctr>\u003Ctd>Evaluation suites\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003Cbr>\u003Csub>50+ metrics\u003C\u002Fsub>\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003Ctd align=\"center\">⚠️\u003Cbr>\u003Csub>Limited\u003C\u002Fsub>\u003C\u002Ftd>\u003C\u002Ftr>\n\u003Ctr>\u003Ctd>\u003Cb>Agent simulation\u003C\u002Fb>\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003Ctd align=\"center\">❌\u003C\u002Ftd>\u003Ctd align=\"center\">❌\u003C\u002Ftd>\u003Ctd align=\"center\">❌\u003C\u002Ftd>\u003Ctd align=\"center\">❌\u003C\u002Ftd>\u003C\u002Ftr>\n\u003Ctr>\u003Ctd>\u003Cb>Voice agent eval\u003C\u002Fb>\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003Ctd align=\"center\">❌\u003C\u002Ftd>\u003Ctd align=\"center\">⚠️\u003Cbr>\u003Csub>Cookbook\u003C\u002Fsub>\u003C\u002Ftd>\u003Ctd align=\"center\">❌\u003C\u002Ftd>\u003Ctd align=\"center\">❌\u003C\u002Ftd>\u003C\u002Ftr>\n\u003Ctr>\u003Ctd>\u003Cb>LLM gateway built in\u003C\u002Fb>\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003Cbr>\u003Csub>100+ providers\u003C\u002Fsub>\u003C\u002Ftd>\u003Ctd align=\"center\">❌\u003C\u002Ftd>\u003Ctd align=\"center\">❌\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003C\u002Ftr>\n\u003Ctr>\u003Ctd>\u003Cb>Guardrails built in\u003C\u002Fb>\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003Cbr>\u003Csub>18 + 15 adapters\u003C\u002Fsub>\u003C\u002Ftd>\u003Ctd align=\"center\">❌\u003C\u002Ftd>\u003Ctd align=\"center\">❌\u003C\u002Ftd>\u003Ctd align=\"center\">❌\u003C\u002Ftd>\u003Ctd align=\"center\">❌\u003C\u002Ftd>\u003C\u002Ftr>\n\u003Ctr>\u003Ctd>\u003Cb>Prompt optimization\u003C\u002Fb>\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003Cbr>\u003Csub>6 algorithms\u003C\u002Fsub>\u003C\u002Ftd>\u003Ctd align=\"center\">❌\u003C\u002Ftd>\u003Ctd align=\"center\">❌\u003C\u002Ftd>\u003Ctd align=\"center\">❌\u003C\u002Ftd>\u003Ctd align=\"center\">❌\u003C\u002Ftd>\u003C\u002Ftr>\n\u003Ctr>\u003Ctd>Prompt management\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003C\u002Ftr>\n\u003Ctr>\u003Ctd>Datasets & experiments\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003C\u002Ftr>\n\u003Ctr>\u003Ctd>No-code eval builder\u003C\u002Ftd>\u003Ctd align=\"center\">✅\u003C\u002Ftd>\u003Ctd align=\"center\">⚠️\u003C\u002Ftd>\u003Ctd align=\"center\">⚠️\u003C\u002Ftd>\u003Ctd align=\"center\">⚠️\u003C\u002Ftd>\u003Ctd align=\"center\">⚠️\u003C\u002Ftd>\u003C\u002Ftr>\n\u003C\u002Ftbody>\n\u003C\u002Ftable>\n\n\u003Csub>Based on publicly-documented features as of April 2026. Corrections welcome — open a PR.\u003C\u002Fsub>\n\n---\n\n## Built for every kind of agent\n\n\u003C!--\n  [MARKETING] use-cases-band.png  (optional — lightweight)\n  What:    Seven small tile icons in a horizontal band: headset (support),\n           phone (voice), briefcase (internal), magnifier (RAG),\n           robot (autonomous), mouse pointer (CUA), code (coding).\n  Size:    1600 × 200, PNG — OR keep as emoji bullets below.\n-->\n\n- **Customer Support:** Ship support AI that customers actually trust\n- **Voice Agents:** Test, evaluate, and improve voice AI end-to-end\n- **Internal Tools:** AI copilots your whole org can rely on\n- **RAG & Search:** Every answer grounded, every citation verified\n- **Autonomous Agents:** Multi-step agents you can actually trust in production\n- **Computer-Use Agents (CUA):** Agents that click with confidence\n- **Coding Agents:** AI that writes code you can actually ship\n\n---\n\n##  Roadmap\n\n[**Vote on the public roadmap →**](https:\u002F\u002Ffutureagi.com\u002Froadmap)  ·  [**GitHub Discussions**](https:\u002F\u002Fgithub.com\u002Forgs\u002Ffuture-agi\u002Fdiscussions\u002Fcategories\u002Froadmap)  ·  [**Releases**](https:\u002F\u002Fgithub.com\u002Ffuture-agi\u002Ffuture-agi\u002Freleases)  ·  [**Changelog**](https:\u002F\u002Fdocs.futureagi.com\u002Fdocs\u002Frelease-notes)\n\n\u003Ctable>\n\u003Ctr>\n\u003Cth width=\"25%\"> Recently shipped\u003C\u002Fth>\n\u003Cth width=\"25%\"> In progress\u003C\u002Fth>\n\u003Cth width=\"25%\"> Coming up\u003C\u002Fth>\n\u003Cth width=\"25%\"> Exploring\u003C\u002Fth>\n\u003C\u002Ftr>\n\u003Ctr valign=\"top\">\n\u003Ctd>\n\n- [x] Prompt optimization engine\n- [x] Taxonomy-based Feed Clustering\n- [x] Agent Runs in Dataset Experiments\n- [x] Simulate from Production Calls\n- [x] LiveKit Configuration via UI\n- [x] System Metric Filtering for Voice\n- [x] Agent Playground\n- [x] Dashboards\n- [x] Access platform via MCP\n- [x] Annotation Queues\n- [x] Command Center (Prism)\n- [x] Open source Future AGI stack\n- [x] Eval Explanation Output Size Control \n\n\n\u003C\u002Ftd>\n\u003Ctd>\n\n- [ ] Agent Changelog & Diff View\n- [ ] Smart Queue Assignment\n- [ ] Essential Node Library for Agent Builder\n- [ ] Full Execution Tracing for Agents\n- [ ] Multi-modal Support for Agents\n\n\u003C\u002Ftd>\n\u003Ctd>\n\n- [ ] Agent Changelog & Diff View\n- [ ] Smart Queue Assignment\n\n\u003C\u002Ftd>\n\u003Ctd>\n\n- [ ] Import agents to Agent Playground\n- [ ] Simulating CUA agents\n- [ ] Simulating Coding agents\n- [ ] Scheduled Simulations\n\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftable>\n\n---\n\n## 🤝 Contributing\n\nWe love contributions — bug fixes, new evaluators, framework integrations, docs, examples, anything.\n\n1.  [Browse `good first issue`](https:\u002F\u002Fgithub.com\u002Ffuture-agi\u002Ffuture-agi\u002Fissues?q=is%3Aissue+is%3Aopen+label%3A%22good+first+issue%22)\n2.  Read the [Contributing Guide](CONTRIBUTING.md)\n3.  Say hi on [Discord](https:\u002F\u002Fdiscord.gg\u002FUjZ2gRT5p) or [Discussions](https:\u002F\u002Fgithub.com\u002Forgs\u002Ffuture-agi\u002Fdiscussions)\n4.  Sign the CLA on your first PR (automatic bot)\n\n\u003C!--\n  [MARKETING] contributors-wall.png  — SKIP UNTIL 50+ CONTRIBUTORS.\n  Stub is intentionally left out. Revisit once contrib count justifies\n  an avatar wall (contrib.rocks auto-generates when enabled).\n-->\n\n---\n\n## 🌍 Community & support\n\n| | |\n|---|---|\n| 💬 [**Discord**](https:\u002F\u002Fdiscord.gg\u002FUjZ2gRT5p) | Real-time help from the team and community |\n| 🗨️ [**GitHub Discussions**](https:\u002F\u002Fgithub.com\u002Forgs\u002Ffuture-agi\u002Fdiscussions) | Ideas, questions, roadmap input |\n| 🐦 [**Twitter \u002F X**](https:\u002F\u002Fx.com\u002FFutureAGI_) | Release announcements |\n| 📝 [**Blog**](https:\u002F\u002Ffutureagi.com\u002Fblog) | Engineering & research posts |\n| 📺 [**YouTube**](https:\u002F\u002Fwww.youtube.com\u002F@Future_AGI) | Walkthroughs & demos |\n| 📊 [**Status**](https:\u002F\u002Fstatus.futureagi.com) | Cloud uptime + incident history |\n| 📧 **support@futureagi.com** | Cloud account \u002F billing |\n| 🔐 **security@futureagi.com** | Private vulnerability disclosure (24h ack — see [SECURITY.md](SECURITY.md)) |\n\n---\n\n##  Telemetry\n\nSelf-hosted Future AGI phones home **anonymous usage counts only** (version, instance ID, feature flags used) so we can size our release testing. **No trace data, no prompts, no API keys**, ever. Opt out via `FUTURE_AGI_TELEMETRY_DISABLED=1`.\n\n---\n\n## ⭐ Star history\n\n\u003Ca href=\"https:\u002F\u002Fstar-history.com\u002F#future-agi\u002Ffuture-agi\">\n  \u003Cpicture>\n    \u003Csource media=\"(prefers-color-scheme: dark)\" srcset=\"https:\u002F\u002Fapi.star-history.com\u002Fsvg?repos=future-agi\u002Ffuture-agi&type=Date&theme=dark\">\n    \u003Cimg alt=\"Star history\" src=\"https:\u002F\u002Fapi.star-history.com\u002Fsvg?repos=future-agi\u002Ffuture-agi&type=Date\">\n  \u003C\u002Fpicture>\n\u003C\u002Fa>\n\n---\n\n## 📄 License\n\nFuture AGI is licensed under the **Apache License 2.0**. See [LICENSE](LICENSE) and [NOTICE](NOTICE).\n\n**You own your evaluation logic and your data.** Inspect every evaluator, every prompt, every trace — no black-box scoring, no vendor lock-in.\n\n---\n\n\u003Cdiv align=\"center\">\n\n**Built with ❤️ by the Future AGI team and [contributors worldwide](https:\u002F\u002Fgithub.com\u002Ffuture-agi\u002Ffuture-agi\u002Fgraphs\u002Fcontributors).**\n\nIf Future AGI helps you ship better AI, a ⭐ helps more teams find us.\n\n[🌐 futureagi.com](https:\u002F\u002Ffutureagi.com) · [📖 docs.futureagi.com](https:\u002F\u002Fdocs.futureagi.com) · [☁️ app.futureagi.com](https:\u002F\u002Fapp.futureagi.com) · [📊 status.futureagi.com](https:\u002F\u002Fstatus.futureagi.com)\n\n\u003C\u002Fdiv>\n","future-agi\u002Ffuture-agi 是一个开源的端到端平台，用于评估、观察和改进大语言模型（LLM）及AI代理应用。其核心功能包括追踪、评估、模拟、数据集管理、网关服务以及安全护栏等，支持从原型设计到生产部署的全生命周期管理。该平台采用Python编写，并且可以自托管，适用于需要对AI系统进行持续优化与监控的企业或研究机构。通过提供统一的反馈循环机制，帮助开发者更快地识别并修复AI代理可能出现的问题，从而提升系统的可靠性和性能。",2,"2026-06-11 02:39:24","CREATED_QUERY"]